WO2008007717A1

WO2008007717A1 - Dynamic image decoding device and dynamic image encoding device

Info

Publication number: WO2008007717A1
Application number: PCT/JP2007/063860
Authority: WO
Inventors: Tomoyuki Yamamoto; Maki Takahashi
Original assignee: Sharp Kabushiki Kaisha
Priority date: 2006-07-11
Filing date: 2007-07-11
Publication date: 2008-01-17

Abstract

A dynamic image decoding device (400a) includes a correction image deciding unit (15) and an inverse filter processing unit (12). For at least one block, the correction image deciding unit (15) performs correction as follows. A pixel value of each of pixels of restoration quantization object image restored by inverse quantization is added by an approximating average pixel value for approximating an average pixel value of a quantization object image in a block to which the pixel belongs. The inverse filter processing unit (12) performs an inverse filter processing equivalent to an inverse conversion of a filter process for removing a frequency component for generating a block distortion. Thus, the dynamic image decoding device (400a) can reduce the block distortion in a decoded dynamic image without causing adverse effect such as blur or flickering of an image. The dynamic image decoding device (400a) can accurately restore a frequency component generating a block distortion removed.

Description

Specification

Moving picture decoding apparatus and moving picture encoding apparatus

Technical field

The present invention relates to a moving picture coding apparatus that divides and encodes a quantization target picture into a plurality of blocks, and a moving picture decoding apparatus corresponding to such a moving picture coding apparatus.

Background art

[0002] In order to efficiently transmit and record a moving image having an enormous amount of information, a technique of dividing image data into a plurality of blocks and quantizing and encoding each block is widely used. ing. In a moving image quantized and encoded for each block, block distortion (block noise) in which pixel values such as luminance levels discontinuously change occurs at the boundary of each block. As a technique for reducing the block distortion, a deblocking filter in the H.264 / AVC moving image encoding system described in Non-Patent Document 1 is known.

[0003] Hereinafter, a moving picture encoding apparatus 100 that encodes a moving picture using the H.264 / AVC moving picture encoding scheme and a moving picture decoding apparatus 200 that decodes a moving picture encoded using the same scheme will be described. This will be described with reference to FIGS.

[0004] (Configuration of Conventional Video Encoding Device)

FIG. 18 is a functional block diagram showing a schematic configuration of a moving picture encoding apparatus 100 that encodes a moving picture by the H.264 / AVC moving picture encoding method. As shown in FIG. 18, the moving image coding apparatus 100 includes a DCT unit 1, a quantization unit 2, a variable length coding unit 3, an inverse quantization unit 4, an I DCT unit 5, a deblocking filter processing unit 6, A frame memory 7, a predicted image derivation unit 8, and an encoding control unit 10 that controls each of the above units are provided.

[0005] The DCT unit 1 divides a difference image obtained by subtracting a prediction image, which will be described later, from an original image into blocks each consisting of 4 × 4 pixels or 8 × 8 pixels, and orthogonally transforms the image signal of each block ( Integer precision DCT). The transform coefficient obtained by the orthogonal transform (corresponding to the DCT coefficient in the discrete cosine transform) is sent to the quantization unit 2. The quantization unit 2 quantizes the transform coefficient of each block according to the quantization parameter supplied from the encoding control unit 10. amount The quantized representative value obtained as a result of the child quantization is sent to the variable length coding unit 3 and the inverse quantization unit 4.

[0006] The variable length encoding unit 3 is based on the various encoding parameters supplied from the encoding control unit 10 and the quantization parameters (quantization representative values supplied from the quantization unit 3). The conversion coefficient of each block) is subjected to variable length coding. The encoded data obtained as a result of encoding by the variable length encoding unit 3 is sent to the moving picture decoding apparatus 200 described later.

[0007] The inverse quantization unit 4 inverts the quantization representative value (transform coefficient of each quantized block) supplied from the quantization unit 3 according to the quantization parameter supplied from the encoding control unit 10. Quantify. That is, the inverse quantization unit 4 restores the transform coefficient of each block from the quantized representative value by the reverse operation of the quantization operation by the quantization unit 2. The transform coefficient of each block restored by inverse quantization is sent to the IDCT unit 5. The IDCT unit 5 converts the transform coefficient of each block obtained by inverse quantization into an image signal in the spatial domain, and restores the difference image. Here, the inverse orthogonal transform applied to the transform coefficient by the IDCT unit 5 is the inverse transform (integer precision IDCT) of the orthogonal transform applied by the DCT unit 1. The local decoded image obtained by adding the difference image restored by the IDCT unit 5 and the predicted image is sent to the deblocking filter processing unit 6.

[0008] The deblocking filter processing unit 6 performs adaptive filtering on the local decoded image in order to remove block distortion in the local decoded image obtained by adding the prediction image and the difference image. Details of adaptive filter processing by the deblocking filter processing unit 6 will be described in detail later.

The locally decoded image from which block distortion has been removed by the deblocking filter processing unit 6 is temporarily stored in the frame memory 7. The frame memory 7 can store a plurality of locally decoded images. The locally decoded image stored in the frame memory 7 is referred to by the predicted image deriving unit 8 as a reference image.

The predicted image deriving unit 8 generates a predicted image from the reference image recorded in the frame memory 7 by performing intra prediction or inter prediction.

[0011] Intra prediction is a process for generating a predicted image by performing intra-frame prediction.

The predicted image derivation unit 8 is defined in the H.264 / AVC video coding standard. Intraframe prediction can be performed in multiple prediction modes (prediction algorithms). When intra prediction is performed, intraframe prediction is performed using the prediction mode specified by the encoding control unit 10.

Inter prediction is a process of generating a prediction image by inter-frame prediction (motion compensation prediction) based on the motion vector determined by the encoding control unit 10 and the reference image stored in the frame memory 7. is there. When performing inter prediction, the predicted image derivation unit 8 uses the block of the size specified by the encoding control unit 10 and also uses inter-frame prediction using a plurality of reference images specified by the encoding control unit 10. I do.

[0013] The encoding control unit 10 determines which prediction method is used to generate a prediction image among intra prediction and inter prediction, and designates the prediction method to the prediction image deriving unit 8. At this time, the encoding control unit 10 determines various encoding parameters according to the prediction method. The coding parameters for performing intra prediction include information for specifying a prediction mode in intra prediction. In addition, the coding parameters for performing inter prediction include information specifying a motion vector, a block size, and a reference image. Furthermore, the encoding control unit 10 designates quantization parameters for the quantization unit 2 and the inverse quantization unit 4.

Next, the encoding operation of moving picture encoding apparatus 100 shown in FIG. 15 will be described. Generally speaking, the moving image encoding apparatus 100 performs encoding of a moving image by repeating the following steps;!

(Step 1) The encoding control unit 10 determines whether intra prediction or inter prediction is performed, and determines an encoding parameter and a quantization parameter necessary for encoding.

(Step 2) According to the determination result in step 1, the predicted image generation unit 8 uses the prediction method specified by the encoding control unit 10 to predict the predicted image based on the reference image stored in the frame memory 7. Is generated.

(Step 3) A difference image between the predicted image generated in Step 2 and the input original image is generated and supplied to the DCT unit 1.

[0018] (Step 4) DCT unit 1, quantization unit 2 and force The image signal of the difference image obtained in step 3 is orthogonally transformed for each block, and the obtained transform coefficient is quantized. The obtained quantization representative value is variable-length encoded by the variable-length encoding unit 3 and output as encoded data. (Step 4) The inverse quantization unit 4, the IDCT unit 5, the force S, and the quantized representative value obtained in step 4 are inversely quantized to restore the difference image.

(Step 5) The difference image restored in Step 4 and the predicted image generated in Step 2 are added, and the obtained locally decoded image is supplied to the deblocking filter processing unit 6.

(Step 6) The deblocking filter processing unit 6 removes block distortion from the local decoded image obtained in step 5, and the local decoded image with reduced block distortion is stored in the frame memory 7 as a reference image. Accumulated.

[0022] (Configuration of Conventional Video Decoding Device)

Next, a moving picture decoding apparatus 200 that decodes a moving picture using the H.264 / AVC moving picture encoding method will be described with reference to FIG.

FIG. 19 is a functional block diagram showing a schematic configuration of a moving picture decoding apparatus 200 that decodes a moving picture by the H.264 / AVC moving picture coding method. As shown in FIG. 19, the moving picture decoding apparatus 200 includes an inverse quantization unit 4, an IDCT unit 5, a deblocking filter processing unit 6, a frame memory 7, a predicted image derivation unit 8, and a variable length decoding unit 20. It has.

Here, among the functional blocks constituting the moving picture decoding apparatus 200, the only functional block not included in the above-described moving picture encoding apparatus 100 (FIG. 18) is the variable length decoding unit 20. Blocks having the same functions as those of the moving picture encoding apparatus 100 have the same names and reference numerals and description thereof is omitted. The variable length decoding unit 20 has a function of variable length decoding the encoding parameter and the quantized representative value (quantized transform coefficient).

In general, the moving picture decoding apparatus 200 shown in FIG. 19 decodes encoded data by repeating the following steps;! To 6.

(Step 1) From the encoded data acquired by the variable length decoding unit 20, the encoding parameter and the quantized representative value (quantized transform coefficient) are variable length decoded.

(Step 2) The predicted image derivation unit 8 generates a predicted image based on the reference image stored in the frame memory 7 according to the decoded encoding parameter.

[0028] (Step 3) Inverse quantization unit 4, IDCT unit 5, force S, quantization representative value obtained in step 1 Is dequantized to restore the difference image.

(Step 4) The difference image restored in Step 3 and the predicted image generated in Step 2 are added, and the obtained decoded image is supplied to the deblocking filter processing unit 6.

(Step 5) The deblocking filter processing unit 6 removes block distortion from the decoded image obtained in step 4, and stores the decoded image with reduced block distortion in the frame memory 7. The decoded image stored in the frame memory 7 can be read at an arbitrary timing and used as a reference image or a display image.

[0031] (Step 6) The decoded image stored in the frame memory is output to an image projection means such as a display at an appropriate timing as a display image.

[0032] (Conventional deblocking filter)

As described above, in the H.264 / AVC moving picture coding system, both the moving picture coding apparatus 100 and the moving picture decoding apparatus 200 use the common deblocking filter processing unit 6! Block distortion that occurs during the quantization / inverse quantization process of moving images is reduced. Hereinafter, the deblocking filter processing unit 6 will be described in a little more detail with reference to FIGS.

[0033] FIG. 20 is an explanatory diagram showing a block division pattern of a quantization target image (difference image between an original image and a predicted image) in a moving image encoding process using the H.264 / AVC moving image encoding method. is there. As shown in FIG. 20, the quantization target image is divided into W X H rectangular blocks. WX H blocks include B ~ in order from the upper left corner of the image to be quantized.

0

B is attached.

W X H- 1

FIG. 21 is a diagram showing two adjacent blocks, block B and block B, in the quantization target image shown in FIG. As shown in FIG. 21, each of block B and block B is composed of a total of 16 pixels arranged in 4 rows and 4 columns. As with block B, all blocks B to B making up the quantization target image in Fig. 20 are arranged in 4 rows and 4 mm.

0 W X H- l n

It consists of 16 pixels arranged.

Each pixel constituting the quantization target image is a combination of a variable n that specifies a block including the pixel and variables u and V that specify the position of the pixel in the block B (n, u, v) Can be specified. That is, the pixel (n, u, v) is a pixel in the u-th column and the v-th row of the block B _n . In the following explanation, the attribute value of the pixel (n, u, v) is expressed as X (n, u, v). For example, the pixel value of the pixel (n, u, v) in the quantization target image is expressed as P (n, u, v).

[0036] Blocks B to B shown in Fig. 20 and Fig. 21 Orthogonalization in video coding

0 WX H- 1

Transformation 'Quantization' Dequantization 'Inverse orthogonal transformation is a processing unit of a series of processing. In other words, the image data of the quantization target image is converted into a conversion coefficient for each of the blocks B to B, and the amount is converted.

0 WX H- 1

It becomes a child. However, since quantization is an irreversible process, even if the image to be quantized is restored by inverse quantization and inverse orthogonal transform, the restored image does not match the image to be encoded. Block distortion occurs at the boundary between adjacent blocks in the restored image. The deblocking filter processing unit 6 reduces this block distortion.

[0037] The deblocking filter processing unit 6 used in the H.264 / AVC moving image coding system includes a filter process between blocks (B, B) adjacent in the horizontal direction and a block (B , B) are performed independently of each other. Deblocky n n + W

The filter strength of the filter processing performed by the filtering filter processing unit 6 is adaptively set according to conditions such as the prediction mode applied to each block.

[0038] As an example of the filter processing performed by the deblocking filter processing unit 6, the filter processing calculation with the strongest filter strength is shown below for blocks B and B adjacent in the horizontal direction. In the following equations, P (n, u, v) is a pixel in the processing target image (local decoded image in the moving image encoding device 100 or decoded image in the moving image decoding device 200) of the deblocking filter processing. The pixel value (n, u, v) and P ′ (n, u, v) represent the pixel value of the pixel (n, u, v) in the filter output image output from the deblocking filter 6.

[0039] country

P '(n, 0, v) = P (nfi, v)

[0040] [Equation 2]

Ρ '(η, Ι, ν) =-(2Ρ («, 0, ν) + 3 尸 (", Ι,) + Ρ (η, 2, ν) + Ρ (η, 3, ν) + Ρ { η + 1,0, ν)} [0041] [Equation 3] η, Ι, ν) + Ρ (η2, ν) + Ρ (η, 3, ν) + Ρ (η + 1,0, ν)

[0042] [Equation 4]

Ρ '(η, 3, ν) = _ (Ρ (η, 1, ν) + 2Ρ (η, 2, ν) + 2 0,3, ν) + 2Ρ (η + 1,0, ν) + Ρ (η + 1,1, ν)

[0043] [Equation 5]

Ρ '0 + 1,0, = upper [Ρ (η, 2, ν) + 2Ρ (η, 3, ν) + 2Ρ (η + 1,0, ν) + 2Ρ (η + 1,1, ν) + Ρ (η + 1,2, ι

[0044] [Equation 6]

Ρ '(η + 1,1, ν) = _ (Ρ {η, Χ ν) + Ρ (η + 1,0, ν) + Ρ (η + 1,1,) + ("+ 1, 2, ι

[0045] [Equation 7]

Ρ '(η + 1,2, ν) =-(P (w, 3, ν) + Ρ {η + 1,0, ν) + Pin + 1,1, ν) + 3Ρ (η + 1,2 , ν) + 2Ρ (η + 1,3, ν)

[0046] [Equation 8]

+ 1,3, ν) = (κ + 1,3, ν)

[0047] Although detailed description is omitted, the same C process is also applied to adjacent blocks in the vertical direction.

[0048] The effects of the filter processing on the processing target image are shown in Figs. 22 (a) to 22 (c).

[0049] (a) of FIG. 22 is a graph showing pixel direct p (n, u, v) of the original image in block B and block B.

FIG. 22 (b) is a graph showing the pixel values P (n, u, v) of the decoded image to be processed by the deblocking filter processing unit 6 in block B and block B. As shown in (b) of FIG. 22, in the decoded image, discontinuous changes in pixel values at block boundaries, that is, block distortion occurs.

[0051] (c) of FIG. 22 is a graph showing pixel values (n, u, v) in block B and block B of the image after filtering the decoded image by the deblocking filter 6 Therefore, discontinuous changes in pixel values in the restored image to be encoded are smoothed, and block distortion is reduced.

Non-Patent Literature: iTU_r Recommendation H.2D4: Advanced Video and oding for generic audiovisual services "(2003)

Disclosure of the invention

[0052] However, in the above-described conventional video encoding device, as a side effect of reducing the block distortion by the deblocking filter, it is different from the block distortion such as image blur or image flicker at the time of moving image reproduction. There was a problem of image quality degradation.

[0053] The side effects of the deblocking filter will be described in more detail below. As is clear from the equations (1) to (8) described above, if the pixel value of the image to be filtered is P (n, u, v), the average pixel value of the block B before the filtering process is P> and the average pixel value <Ρ '> of block B after filtering are given by the following equations, respectively.

[0054] [Equation 9] ヽ^{η 1} 〉 = ^∑ ∑ ^Ρ ( ^Η

1 ⁰ ι, = 0 V = 0

[0055] [Equation 10]

That is, the average pixel value in each block is different before and after the filter processing. Therefore, when the conventional deblocking filter is used, the average pixel value in each block is not saved before and after the filter processing! /, So the average pixel value between the encoding target image and the decoded image is obtained by the filter processing. The difference can be enlarged.

[0057] Then, the prediction image is generated in the moving image encoding apparatus using the local decoded image having a difference in average pixel values as compared to the encoding target image as a reference image. In addition, in the video decoding device, a decoded image having an average pixel value difference with respect to the encoding target image is generated as a reference image or a display image. For this reason, blurring or flickering of the image in the moving image occurs. [0058] In order to solve such a problem, the applicant applies a filtering process in Japanese Patent Application "Japanese Patent Application No. 2006-0756 85" to remove frequency components that generate block distortion from an encoding target image. A moving picture encoding apparatus provided with filter processing means and a moving picture decoding apparatus provided with inverse filter processing means corresponding to the inverse transformation of the filter processing have been proposed. Thus, the frequency component that generates block distortion is removed from the encoding target image by the filter processing unit, and the frequency component removed by the inverse filter processing unit is restored in the decoded image. it can. Moreover, the filter processing means can be configured to keep the average value of the pixel values unchanged before and after the filter processing. Accordingly, it is possible to provide a moving image encoding apparatus and a moving image decoding apparatus that can reduce block distortion in a decoded moving image and that do not cause side effects such as blurring and flickering of the image.

[0059] However, since the quantization process is an irreversible process, the output image of the filter processing means and the input image of the inverse filter processing means have a discrepancy at least for the quantization error. Therefore, the frequency component restored by the inverse filter processing unit cannot accurately reproduce the frequency component removed from the encoding target image by the filter processing unit.

[0060] The present invention has been made in view of the above-described problems, and an object of the present invention is to reduce block distortion in a decoded moving image and to prevent side effects such as image blur and flicker. It is a moving image encoding device and a moving image decoding device, and is to realize a moving image encoding device and a moving image decoding device that can restore more accurately the frequency components that generate the removed block distortion. .

[0061] In order to solve the above-described problem, the moving image encoding apparatus of the present invention quantizes a quantization target image subjected to a filter process for removing a frequency component that generates block distortion for each block. A video decoding device that inverse-quantizes the quantized representative value obtained by the above, and for at least one block, the restoration value restored by the inverse quantization The pixel value of each pixel of the quantization target image A restored quantization target image correcting means for performing correction by adding an approximate average pixel value approximating the average pixel value of the quantization target image in the block to which the pixel belongs, and the restored quantization target image subjected to the correction In contrast to the above And an inverse filter processing means for performing an inverse filter process corresponding to an inverse transformation of the filter process.

[0062] In a moving image encoding apparatus that divides a quantization target image subjected to filter processing to remove a frequency component that generates block distortion into a plurality of blocks, and quantizes the quantization target image for each block. The average pixel value of each block of the quantization target image is preferably as small as possible. This is because by reducing the average pixel value of the quantization target image, the quantization level interval in the quantization can be set small, and the quantization error caused by the quantization can be reduced. is there.

[0063] For this reason, in the moving image encoding apparatus, the approximate average pixel value that approximates the average value of the quantization target image in the block to which the pixel belongs is subtracted from the pixel value of each pixel of the quantization target image. It is preferable to perform correction. This is because the average pixel value of each block of the corrected quantization target image can be reduced by performing the above correction. By using the approximate average pixel value that better approximates the average pixel value of the quantization target image, the average pixel value of each block of the corrected quantization target image is further reduced by the force S.

[0064] According to the above configuration, the same approximate average pixel value that is subtracted from the quantization target image by the moving image encoding device is used as the restored quantization target image by the moving image decoding device. Can be added. For this reason, the difference between the quantization target image after the filtering process and the restored quantization target image before the inverse filtering process is determined as only the quantization error that occurs during the quantization process. It becomes possible to do.

In addition, the average pixel value of each block of the quantization target image is gradually reduced by the correction performed on the moving image encoding device side, so that it is generated in the quantization process. The amount of quantization error to be made is smaller than when the above correction is not performed. Then, there is a difference between the quantization target image after the filtering process is performed by the moving image encoding apparatus and the restored quantization target image before the inverse filtering process is performed by the moving image decoding apparatus. This is smaller than when the above correction is not performed. Therefore, the inverse filter processing means can restore the frequency component removed by the filter processing means more accurately than when the correction is not performed. [0066] That is, according to the above configuration, there is an effect that it is possible to decode an image in which an image encoded by the moving image encoding apparatus is reproduced more loyally.

[0067] Note that the quantization target image may be the encoding target image itself encoded by the moving image encoding device, or a difference obtained by subtracting a predicted image from the encoding target image. It may be an image. Here, examples of the predicted image include a predicted image generated by using an intra-frame prediction process, an inter-frame prediction process, or the like using a locally decoded image as a reference image.

[0068] Further, the inverse filter processing performed by the inverse filter processing means may correspond to a strict inverse transformation or an approximate inverse transformation with respect to the filter processing of the filter processing means. It may be what you do. In other words, the inverse filtering process is sufficient if it approximates a strict inverse transform within a range that does not adversely affect the quality of the final moving image. For example, the calculation accuracy of the filtering process (for example, integer precision) A degree of error is acceptable.

[0069] In the moving picture decoding apparatus according to the present invention, correction block designation information acquisition means for acquiring correction block designation information for designating a block to be corrected from the moving picture encoding apparatus for performing the quantization. In addition, it is preferable that the restored quantization target image correcting unit performs the above correction on a block specified by the correction block specifying information.

[0070] According to the above configuration, the video decoding device acquires the correction block designation information from the video coding device, and performs the correction on the block designated by the correction block designation information. . As a result, the moving picture coding apparatus designates the corrected block designation information as a block that has been corrected to subtract the approximate average pixel value from the quantization target image, so that the moving picture decoding apparatus It is possible to perform correction by adding the approximate average value to the same block of the image to be converted.

[0071] For this reason, it is possible to match the quantization target image after the filtering process and the restored quantization target image before the inverse filtering process more accurately.

[0072] Note that the correction block designation information may be encoded and provided from the moving picture coding apparatus, for example. In this case, the correction block designation information acquisition procedure The stage can be configured to decode the encoded data and obtain the correction block designation information.

[0073] The moving picture decoding apparatus according to the present invention includes an average pixel value before correction obtained by averaging pixel values of each block of the restored quantization target image, and the approximate average pixel of the block from the average pixel value before correction. A comparison estimation unit that estimates a magnitude relationship with a corrected average pixel value obtained by subtracting a value, and the restored quantization target image correction unit uses the comparison estimation unit to change the corrected average pixel value to the average pixel before correction. It is preferable that the correction is performed for blocks estimated to be smaller than the value.

[0074] According to the above configuration, the comparative estimation means estimates the magnitude relationship between the magnitude of the average pixel value before correction and the magnitude of the average pixel value after correction, and determines the magnitude of the average pixel value after correction. For the block estimated to be smaller, the correction is performed by the restored quantization target image correcting means. Therefore, the moving picture coding apparatus that performs the quantization performs correction for subtracting the approximate average pixel value for a block in which the size of the average pixel value after correction is smaller than the size of the average pixel value before correction. As the accuracy of the estimation increases, the quantization target image after the filtering process and the restored quantization target image before the inverse filtering process can be matched more accurately. Become possible

[0075] Moreover, according to the above configuration, it is not necessary to receive information from the moving image encoding device that performs the quantization in order to specify the block to be corrected. Therefore, there is a further effect that the amount of code can be reduced, and the power and burden of decoding processing can be reduced.

Note that the comparison estimation means estimates, for example, that the average pixel value before correction matches the average pixel value after correction for a block in which the value of the approximate average pixel value is zero. When the approximate average pixel value is larger than a predetermined threshold value, it can be estimated that the corrected average pixel value is larger than the pre-correction average pixel value.

[0077] In the video decoding apparatus according to the present invention, it is preferable that the approximate average pixel value is derived from the restored quantization target image! /.

[0078] According to the above configuration, the approximate average pixel value is derived from the restoration quantization target. Therefore, the moving image decoding apparatus performs the approximation without referring to the quantization target image. An average pixel value can be derived. On the other hand, the moving image coding apparatus that performs the quantization also performs the inverse quantization on the quantization representative value obtained by the quantization so that the same image as the restored quantization target image is stored in the local apparatus. And the approximate average pixel value can be derived. Therefore, the moving picture decoding apparatus and the moving picture encoding apparatus can share the same approximate average pixel value. For this reason, there is an additional effect that the quantization target image after the filtering process and the restored quantization target image before the inverse filtering process can be matched more accurately. .

[0079] The video decoding device according to the present invention converts the approximate average pixel value of each block to the pixel value of the restored quantization target image in the adjacent block for which the approximate average pixel value is derived adjacent to the block. An approximate average pixel value deriving unit for deriving based on the above, and the approximate average pixel value deriving unit for the adjacent block! /, The correction is not performed! / When the average pixel value of the quantization target image is the approximate average pixel value of the block and the correction is performed for the adjacent block, the average pixel value of the restored quantization target image in the adjacent block and the adjacent pixel It is preferable that the difference value from the approximate average pixel value of the block is the approximate average pixel value of the block.

[0080] According to the above configuration, the approximate average pixel value that closely approximates the quantization target image can be obtained based on the restored quantization target image. For this reason, it is possible to more accurately match the quantization target image after the filtering process and the restored quantization target image before the inverse filtering process. Play.

Note that the video decoding device may derive the approximate average pixel value based on the pixel value of the restored quantization target image in one adjacent block! /, The approximate average pixel value may be derived based on the pixel values of the restored quantization target image in a plurality of adjacent blocks.

Further, the average pixel value in a certain block of the restored quantization target image is obtained by, for example, inverse quantization of the restored quantization target image in the frequency domain in the block (the quantization representative value). It is possible to calculate from the DC component (corresponding to the DCT coefficient). [0083] The moving picture decoding apparatus according to the present invention provides an approximation that correlates the approximate average pixel value of each block with the block selected by referring to a quantization parameter that specifies the quantization method. Approximate average pixel value deriving means for deriving based on the pixel value of the restored quantization target image in the correlation block for which the average pixel value has already been derived, wherein the approximate average pixel value deriving means performs the correction for the correlation block. If the correction is performed on the correlation block when the average pixel value of the restored quantization target image in the correlation block is set as the approximate average pixel value of the block, the restoration in the correlation block is performed. The difference value between the average pixel value of the quantization target image and the approximate average pixel value of the correlation block is used as the approximate average pixel value of the block. Arbitrariness is preferred.

[0084] According to the above configuration, it is possible to obtain the approximate average pixel value that closely approximates the quantization target image. For this reason, it is possible to further accurately match the quantization target image after the filtering process and the restored quantization target image before the inverse filtering process. Play.

Note that the moving image decoding apparatus can accurately derive the approximate average pixel value of each block with reference to, for example, a prediction mode used for intra prediction. More specifically, a block whose prediction mode matches the prediction mode in the block is selected from among the blocks close to each block, and approximated as the average pixel value of the restored quantization target image in the selected block. An average pixel value can be derived.

[0086] In order to solve the above-described problem, a video encoding device according to the present invention is a video encoding device that divides a quantization target image into a plurality of blocks and quantizes each block. Filter processing means for performing a filtering process on the quantization target image to remove a frequency component that generates block distortion, and each of the quantization target images on which at least one block has been subjected to the finer processing. Quantization target image correction means for performing correction by subtracting an approximate average pixel value that approximates the average pixel value of the quantization target image in the block to which the pixel belongs from the pixel value of the pixel; It is characterized by.

[0087] The filter processing means removes a frequency component that generates block distortion from the encoding target image before quantization. Therefore, the quantization target image to be quantized above The frequency components that generate block distortion are removed in advance. Therefore, the restored quantization target image obtained by quantizing and dequantizing the quantization target image subjected to the filtering process by the filter processing means has reduced block distortion. Moreover, the frequency component previously removed by the filter processing means can be restored by inverse filter processing corresponding to the inverse transformation of the filter processing of the filter processing means. In other words, the moving picture coding apparatus having the above-described configuration can generate coded data that can restore a decoded image with reduced block distortion without losing a specific frequency component. And! /, Has the effect.

According to the above configuration, the average power of the quantization target image in the block to which the pixel belongs is calculated from the pixel value of each pixel of the quantization target image for at least one block. Since correction for subtracting the approximate average pixel value that approximates the value is performed, it is possible to reduce the average pixel value of each block of the quantization target image below the original average pixel value of the block. Therefore, the quantization level interval in the quantization can be set small, and the quantization error caused by the quantization can be reduced. For this reason, the quantization error generated in the quantization process can be further reduced. Therefore, the quantization target image after the filtering process can be more accurately matched with the restored quantization target image before the inverse filtering process.

[0089] For this reason, in the inverse quantization, it is possible to restore the frequency component removed by the filtering process more accurately. In other words, it is possible to generate encoded data that can restore a decoded image that faithfully reproduces the encoding target image.

[0090] The moving image encoding apparatus according to the present invention includes, for each block of the quantization target image, an average pixel value before correction obtained by averaging pixel values of the quantization target image in the block, and the average pixel before correction. Average pixel value calculating means for calculating a corrected average pixel value obtained by subtracting the approximate average pixel value of the block from the value, and the average pixel value before correction for each block of the quantization target image! And a correction evaluation unit that compares the average pixel value after correction with the magnitude of the average pixel value after correction. The quantization target image correction unit includes a correction evaluation unit that causes the average pixel value after correction to be greater than the average pixel value before correction. The block determined to be small It is preferable that the above correction is performed for the lock.

[0091] According to the above configuration, the average pixel value before correction and the average pixel value after correction are compared in magnitude, and the block for which the average pixel value after correction is smaller than the average pixel value before correction Correction is performed. For this reason, the average pixel value of each block of the quantization target image can be reliably made equal to or less than the original average pixel value of the block. Therefore, the quantization level interval in the above quantization can be set smaller, and the quantization error generated by the quantization can be reduced. For this reason, it is possible to more accurately match the quantization target image after the filtering process and the restored quantization target image before the inverse filtering process. That is, there is an additional effect that encoded data that can restore a decoded image that more faithfully reproduces the encoding target image can be generated.

[0092] In the moving image encoding device according to the present invention, the quantization target image correcting unit corrects the moving image decoding device that inversely quantizes the quantized representative value obtained by the quantization. It is preferable that a correction block designation information providing unit for providing correction block designation information for designating a block is further provided.

[0093] According to the above configuration, it is possible to provide the moving picture decoding apparatus with corrected block designation information for designating the block subjected to the correction in the moving picture encoding apparatus. For this reason, the video decoding device can perform correction complementary to the correction on the same block as the block on which the video encoding device has performed the correction based on the correction block designation information. Become. For this reason, it is possible to match the quantization target image after the filtering process and the restored quantization target image before the inverse filtering process more accurately. That is, there is an additional effect that encoded data that can restore a decoded image that more accurately reproduces the encoding target image can be generated.

[0094] In order to solve the above-described problem, the moving picture decoding apparatus according to the present invention quantizes a quantization target image subjected to a filter process for removing a frequency component that generates block distortion for each block. A video decoding device that inversely quantizes the obtained quantized representative value, wherein at least one block is restored by the above inverse quantization A restored quantization target image correction unit that performs correction by subtracting an approximate quantization error that approximates a quantization error in a block to which the pixel belongs from the pixel value of each pixel of the quantization target image; And an inverse filter processing means for performing an inverse filter process corresponding to the inverse transform of the filter process on the restored quantization target image.

[0095] The moving picture decoding apparatus quantizes the quantization target image that has been subjected to the filter processing for removing the frequency component that generates block distortion. Here, since the quantization target image is obtained by removing the frequency component that generates the block distortion in advance, the block distortion in the restored quantization target image restored by inverse quantization is suppressed. The restored quantized image is restored by the inverse filter processing means with the frequency component removed during the encoding process. Therefore, the restored quantization target image that has been subjected to the above inverse filter processing is an image that has a reduced block distortion and does not lack a specific frequency component, that is, closely approximates the quantization target image. It becomes an image.

In addition, the moving picture decoding apparatus includes a decoded image correcting unit that subtracts an approximate quantization error that approximates a quantification error from the restored quantization target image, and is restored to the inverse filter processing means. The quantization error in the quantization target image is reduced. For this reason, the restored quantization target image that is the processing target of the above inverse filtering process is a better approximation of the quantization target image that has been subjected to the filtering process. Therefore, the inverse filter processing means can more accurately restore the frequency component removed during the encoding process. That is, according to the above configuration, it is possible to obtain a decoded image in which the encoding target image is reproduced more faithfully.

[0097] The moving picture decoding apparatus according to the present invention obtains correction block designation information acquisition means for acquiring correction block designation information for designating a block to be corrected from the moving picture encoding apparatus that performs the quantization. It is preferable that the restored quantization target image correcting unit performs the correction on the block specified by the correction block specifying information.

[0098] According to the above configuration, the moving picture decoding apparatus acquires the correction block designation information from the moving picture encoding apparatus, and assigns the corrected block designation information to the block designated by the correction block designation information. The above correction is performed. As a result, the moving picture encoding apparatus can designate the block to be corrected to the moving picture encoding apparatus.

[0099] Note that the correction block designation information may be encoded and provided from the moving picture coding apparatus, for example. In this case, the correction block designation information acquisition means can be configured to decode the encoded data and acquire the correction block designation information.

[0100] In the video decoding device according to the present invention, the approximate quantization error in each block is a weighted average value obtained by weighted average of the average pixel values of the restored quantization target image in each of the blocks in the vicinity of the block. And the difference value from the average pixel value of the restored quantization target image in the block!

[0101] The weighted average value is a good approximation of the restored quantization target image having no quantization error in each block. That is, the approximate quantization error obtained by subtracting the weighted average value from the average pixel value of the restored quantization target image is a good approximation of the true quantization error. Therefore, according to the above configuration, there is a further effect that it is possible to effectively reduce the quantization error in the restored quantization target image passed to the inverse filter processing means.

[0102] In order to solve the above-described problem, the moving picture encoding apparatus according to the present invention provides a plurality of encoding target images to which a filtering process for removing a frequency component that generates block distortion is performed by a filtering process. A moving picture coding apparatus that divides into blocks and codes each block, local decoding means for decoding the coded data obtained by the above coding to obtain local decoded images, and for each block, Any one of the weighted average value obtained by weighted averaging the average pixel value of the local decoded image in each block adjacent to the block and the average pixel value of the local decoded image in the block is the value in the block. The correction method determining means for determining whether the image is close to the average pixel value of the encoding target image and the moving image decoding apparatus for decoding the encoded data obtained by the encoding described above. Correction block instruction information that provides correction block instruction information for indicating a block that is determined to have a value that is closer to the average pixel value of the image to be encoded! / As a block that should be corrected. Providing means. [0103] According to the above configuration, for each block! /, A weighted average value obtained by weighted average of the average pixel values of the local decoded image in each of the blocks in the vicinity of the block, and in the block It is determined which of the average pixel values of the local decoded image is close to the average pixel value to be encoded in the block. Then, with respect to the moving picture decoding apparatus that decodes the encoded data obtained by the encoding, the weighted average value is determined to be closer to the average pixel value of the encoding target image! / Then, correction block instruction information for instructing to correct the decoded image is provided.

[0104] As an example, the moving picture decoding apparatus performs, for example, correction for subtracting the approximate quantization error in the block to which the pixel belongs from the pixel value of each pixel of the decoded image for at least one or more blocks. If the approximate quantization error in each block is, for example, a value obtained by subtracting the weighted average value from the average pixel value of the decoded image in the block, the video decoding device The quantization error can be effectively reduced by the correction based on the correction block instruction information.

[0105] Other objects, features, and advantages of the present invention will be sufficiently understood from the following description. The advantages of the present invention will be apparent from the following description with reference to the accompanying drawings.

Brief Description of Drawings

FIG. 1 is a functional block diagram showing a configuration of a video decoding device according to a first embodiment of the present invention.

FIG. 2 is a functional block diagram showing a basic form of a moving picture coding apparatus according to an embodiment of the present invention.

FIG. 3 is a functional block diagram showing a basic form of a video decoding apparatus according to an embodiment of the present invention.

FIG. 4 is a flowchart showing an outline of filter processing by a filter processing unit included in the video encoding device according to the embodiment of the present invention.

FIG. 5 is a graph showing the effect of the filter processing shown in FIG. (A) is a graph showing pixel values of the processing target image to be filtered, (b) is a graph showing average pixel values for each block of the processing target image, and (c) is a graph. Lines using average pixel values shown in (b) (D) is the pixel value of the filter output image.

FIG. 6 is a flowchart showing an overview of inverse filter processing by an inverse filter processing unit included in a video encoding device and video decoding device according to an embodiment of the present invention.

FIG. 7 is a graph showing the operation of the inverse filter processing shown in FIG. (A) is a graph showing the pixel value of the processing target image to be subjected to the inverse filter processing, (b) is a graph showing the average pixel value for each block of the processing target image, and (c) is a graph. (B) is a graph showing the predicted value obtained by linear interpolation using the average pixel value shown in (b), and (d) is a graph showing the pixel value of the inverse filter output image.

FIG. 8 is a functional block diagram showing a configuration of a video encoding device according to the first embodiment of the present invention.

FIG. 9 shows an embodiment of the present invention, and is a functional block diagram showing a configuration of a correction determination unit in the moving picture encoding apparatus shown in FIG.

10] An embodiment of the present invention, showing a configuration of a moving image encoding device shown in FIG. 8 and a corrected image candidate deriving unit in the moving image decoding device shown in FIG. FIG.

11] A flowchart showing an embodiment of the present invention, and showing a flow of a corrected image candidate derivation process in the moving picture encoding apparatus shown in FIG. 8 and the moving picture decoding apparatus shown in FIG. It is.

FIG. 12 is a functional block diagram showing a configuration of a video encoding device according to a second embodiment of the present invention.

FIG. 13 is a functional block diagram showing a configuration of a video decoding device according to a second embodiment of the present invention.

FIG. 14 is a flowchart showing a flow of correction method determination processing in the video encoding apparatus shown in FIG. 12.

FIG. 15 is a graph showing the operation of the moving picture encoding apparatus and moving picture decoding apparatus according to the second embodiment of the present invention, and showing the pixel value of each pixel.

16] An embodiment of the present invention is shown, and the moving picture encoding device shown in FIG. 14 is a flowchart showing a flow of inverse filter input correction processing in the video decoding apparatus shown in FIG.

FIG. 17 is a graph for explaining distortion in a decoded image caused by a quantization error in the video encoding device shown in FIG. 2 and the video decoding device shown in FIG. 3. FIG. 18 is a functional block diagram showing a conventional technique and showing a configuration of a moving image encoding device including a deblocking filter.

[Fig. 19] Fig. 19 is a functional block diagram showing a conventional technique and showing a configuration of a video decoding device corresponding to the video encoding device shown in Fig. 18.

FIG. 20 is an explanatory diagram showing an image division pattern in a quantization target image to be quantized or a processing target image to be filtered.

21] An enlarged view showing two adjacent blocks in an image divided into a plurality of blocks according to the division pattern shown in FIG.

Sono 22] This is a graph showing the prior art and showing the action of the deblocking filter. (A) is a graph showing pixel values of an encoding target image, (b) is a graph showing pixel values of a locally decoded image including block distortion, and (c) is a deblocking filter output image. It is a graph which shows a pixel straight.

Explanation of symbols

300, 300a, 300b video encoding device

400, 400a, 400b video decoding device

1 DCT section

2 Quantization part

3 Variable length encoder (Correction block designation information providing means)

4 Inverse quantization section (inverse quantization means)

5 IDCT section

6 Deblocking filter processor

7 frame memory

8 Predictive image deriving unit

10 Coding control unit 11 Filter processing section (filter processing means)

12 Inverse filter processing unit (inverse filter processing means)

13 Correction determination unit (average pixel value calculation means, correction evaluation means)

14 Corrected image candidate derivation unit (approximate average pixel value derivation means)

15 Correction image determination unit (quantization target image correction means,

Restoration quantization target image correction means)

16 Inverse filter input correction unit (decoded image correction means)

17 Correction method decision unit (Correction method judgment means)

20, 20a, 20b Variable length decoder (correction block designation information acquisition means)

21 Coding control unit

50 Average pixel value calculation unit (mean pixel value calculation means)

51 Corrected image candidate evaluation section (correction evaluation means)

52 Prediction error recording memory

53 Correction amount deriving section

54 Corrected image candidate generator

55 Prediction error derivation unit

BEST MODE FOR CARRYING OUT THE INVENTION

[Basic configuration]

First, the basic configuration of the video encoding device and the video decoding device will be described with reference to FIGS. Then, after pointing out the problems in the basic configuration, the video encoding device and the video decoding device according to the present embodiment realized by modifying the basic configuration will be described. .

[0109] (Basic configuration of video encoding device)

FIG. 2 is a functional block diagram showing a main configuration of the moving picture coding apparatus 300. As shown in FIG. 1, the moving picture coding apparatus 300 includes a DCT unit 1, a quantization unit 2, a variable length coding unit 3, an inverse quantization unit 4, an IDCT unit 5, a frame memory 7, and a predicted image derivation unit. 8, an encoding control unit 21, a filter processing unit 11, and an inverse filter processing unit 12 are provided.

[0110] Main Phases of the Video Encoding Device 300 in Fig. 2 and the Conventional Video Encoding Device 100 (Fig. 18) The difference is that the moving picture coding apparatus 300 includes a filter processing unit 11 and an inverse filter processing unit 12 instead of the conventional deblocking filter 6, and instead of the conventional coding control unit 10. In addition, the encoding control unit 21 is provided. In FIG. 2, blocks having the same functions as those of the conventional moving image encoding apparatus 100 are denoted by the same names and reference numerals as those in FIG. 18, and description thereof is omitted.

[0111] The functions of the encoding control unit 21, the filter processing unit 11, and the inverse filter processing unit 12, which are characteristic configurations of the moving image encoding device 300, are described as follows.

[0112] In addition to the various functions of the encoding control unit 10 in the moving image encoding device 100 (Fig. 18), the encoding control unit 21 instructs the inverse filter processing unit 12 to execute the inverse filter processing. Has a function of transmitting a control signal. The control signal is transmitted every predetermined encoding processing unit (for example, every block or every frame).

[0113] The filter processing unit 11 performs filter processing to remove a specific frequency component FL that generates block distortion from the processing target image. The processing target image to be processed by the filter processing unit 11 is an encoding target image. In the DCT unit 1, a differential image obtained by subtracting the prediction image generated by the prediction image deriving unit 8 from the encoding target image (filter output image) subjected to the filter processing by the filter processing unit 11 is quantized. Supplied as a target image.

On the other hand, the inverse filter processing unit 12 acquires a locally decoded image recorded in the frame memory 7 at the timing indicated by the control signal supplied from the encoding control unit 21. The inverse filter processing unit 12 uses the locally decoded image as a processing target image, and performs an inverse filtering process corresponding to the inverse transformation of the filter processing performed by the filter processing unit 11 on the processing target image. The locally decoded image (inverse filter output image) that has been subjected to the inverse filter processing by the inverse filter processing unit 12 is stored in the frame memory 7 as a reference image, and is used by the predicted image deriving unit 8 to generate the next predicted image. .

[0115] The DCT unit 1 and the quantization unit 2 divide the quantization target image into a plurality of blocks and quantize the image data of each block. Accordingly, block distortion may be included in a decoded image or a local decoded image obtained by adding a predicted image to a restored quantization target image restored by inverse quantization of the quantization representative value obtained by the quantization. That is, a locally decoded image generated inside the video encoding device 300 to generate a predicted image, and a video to be described later. The decoded image generated by the image decoding device 400 may include block distortion. However, in the moving picture coding apparatus 300, the quantization target image supplied to the DCT unit 1 is a difference image between the filter output image subjected to the filter processing by the filter processing unit 11 and the predicted image. Here, since the filter processing unit 11 acts on the processing target image so as to remove the frequency component FL that generates block distortion, it is possible to effectively reduce block distortion generated in the process of quantization and dequantization. .

In other words, the spatial frequency component lost in the quantization process in the conventional video encoding device 100 is temporarily removed by the filter processing unit 11 in the video encoding device 300, and the inverse filter is used. By restoring by the processing unit 12, it is possible to avoid the influence of the quantization process and suppress the occurrence of block distortion.

[0117] Also, the filter processing unit 11 and the inverse filter processing unit 12 are provided with a conventional moving picture encoding device.

Unlike the deblocking filter processing unit 6 in 100, as described later, it can be configured to maintain the average pixel value (for example, average luminance level) before and after processing in each block. Therefore, it is possible to avoid problems such as image blurring and flickering during moving image playback caused by the deblocking filter processing.

[0118] (Basic configuration of video decoding device)

FIG. 3 is a functional block diagram showing a schematic configuration of a moving picture decoding apparatus 400 corresponding to the moving picture encoding apparatus 300 shown in FIG. As shown in FIG. 3, the video decoding device 400 includes an inverse quantization unit 4, an IDCT unit 5, a frame memory 7, a predicted image derivation unit 8, a variable length decoding unit 20, and an inverse filter processing unit 12. ing.

[0119] The difference between the moving picture decoding apparatus 400 of Fig. 3 and the conventional moving picture decoding apparatus 200 (Fig. 19) is that the moving picture decoding apparatus 400 uses an inverse filter processing unit instead of the conventional deblocking filter 6. It is a point with 12. In FIG. 3, blocks having the same functions as those of the conventional video decoding device 200 are denoted by the same names and symbols as in FIG. 19, and the description thereof is omitted.

[0120] The encoded data decoded by the video decoding device 400 is generated by the video encoding device 300 based on the image from which the frequency component FL that generates block distortion is removed! /. is there. The variable length decoding unit 20 of the video decoding device 400 decodes this encoded data. To do. The inverse quantization unit 4 and the IDCT unit 5 perform inverse quantization on the value obtained by the decoding. The restored quantization target image restored by inverse quantization corresponds to a difference image obtained by subtracting the predicted image from the original image. The moving picture decoding apparatus 400 generates a decoded image by adding the predicted image generated by the predicted image deriving unit 8 to the restored difference image.

[0121] The inverse filter processing unit 12 included in the video decoding device 400 performs the same inverse filter processing as the inverse filter processing unit 12 included in the video encoding device 300. That is, the inverse filter processing unit 12 acts on the decoded image lacking the frequency component FL that generates block distortion to restore the frequency component FL removed by the filter processing. The inverse filter output image that has been subjected to the inverse filter processing by the inverse filter processing unit 12 is referred to by the prediction image deriving unit 8 as a reference image for generating a prediction image, and is output to a display or the like as a reproduction image. The

[0122] As described above, the frequency component FL removed from the encoding target image by the operation of the inverse filter processing unit 12 is restored in the reproduced image or the reference image. That is, the moving picture decoding apparatus 400 can output a reproduced picture in which a specific frequency component is not lost at the same time that the block distortion is reduced. In addition, the predicted image can be generated by the predicted image deriving unit 8 based on a reference image in which specific frequency components are not lost at the same time that block distortion is reduced.

[0123] (Details of filtering)

Next, filter processing performed by the filter processing unit 11 included in the moving image encoding device 300 will be described in detail.

[0124] The filter processing unit 11 performs horizontal filter processing and vertical filter processing by dividing a processing target image to be filtered into a plurality of blocks. Here, the division pattern in which the filter processing unit 11 divides the processing target image is as shown in FIGS. 20 and 21, and the DCT unit 1 and the quantization unit 2 perform the quantization target image for quantization. This is the same as the division pattern for dividing.

[0125] The filter processing unit 11 uses the image data of the block B in the filter output image output from the filter processing unit 11 as the image data of the block B in the processing target image. It is calculated from the image data of the block adjacent to the block. In horizontal finisher processing, the image data of block B is calculated with reference to the image data of block B adjacent to the right side of the block B. In vertical filtering, the image data of block B is calculated. Is the image data n n + W of block B adjacent to the lower side of the block.

Calculated with reference to the data.

Hereinafter, the filter processing calculation executed by the filter processing unit 11 in the horizontal filter processing will be described with reference to FIG. 4 and FIG. In the following description, the force filter processing unit 11 for describing the filter processing calculation for calculating the pixel value (for example, the luminance level) of one block B performs the filter processing calculation described below for all adjacent two blocks. Repeating the lock completes the horizontal filtering process. Note that the above repetitions are the odd and odd numbers adjacent to each other, such as (B, B), (B, B), (B, B)

1 2 3 4 5 6

A pair of several blocks may be paired, and the above filtering processing may be executed for all of these pairs. (B, B), (B, B), (B, B) So that all the blocks

1 2 2 3 3 4

The above filtering processing calculation may be executed with reference to the next block. However, in these repetitions, a pair of non-adjacent blocks, for example, block B (k is an integer) at the right end of the image and the next program k XW-1 located at the lower left end of the block

The above filter processing is not performed for the pair consisting of ACK B.

k XW

FIG. 4 is a flowchart for schematically explaining the filter processing calculation executed by the filter processing unit 11. As shown in FIG. 4, the filter processing calculation by the filter processing unit 11 includes step S1 for calculating the average pixel value, step S2 for calculating the predicted value, and step S3 for calculating the filter output image. It is out.

[0128] Steps S1 to S3 will be described in more detail as follows. In the following description, the filter processing operation for calculating the pixel value of the 4 pixels arranged in the Vth row out of the 16 pixels arranged in 4 rows and 4 columns belonging to the block B will be described as! / . The filter processing unit 1 1 performs the filtering processing described below on the first row force of block B and the fourth row in order or in parallel, so that all the pixels belonging to block B are processed. Calculate the value.

[0129] (Step S1) The filter processing unit 11 performs processing on the Vth row of the block B in the processing target image. The average pixel value p> of the four pixels arranged in the self-sequence and the average pixel value p> of the four pixels arranged in the Vth row of the block B _{+ ι} in the processing target image are calculated. Filter processing unit 11 n + l,

The calculation formula for calculating the force s and the average pixel value is as follows.

[0130] [Equation 11]

Here, p (n, u, v) is a pixel value in the pixel (n, u, v) of the processing target image. (A) of FIG. 5 is a graph showing pixel values of the processing target image. (B) of FIG. 5 is a graph showing the average pixel value obtained by! / In step S1.

[0132] (Step S2) The filter processing unit 11 determines that the average pixel value obtained in Step S 1 <p

,

> And <>, the predicted value P (n, u, V) for each pixel in the V row of block B and the predicted value p ( n + 1 pred n + 1 pred

, U, v). The filter processing unit 11 calculates the predicted values p (1, 11,) and (n + l, u, v) pred pred

The calculation formula for calculating is as follows.

[0133] [ _Equation 12] p _pred (n, u, v) =-(〈〉-„ ₊₁ ,) '+ (11 〈/〉-3 Reno)

[0134] [Equation 13] ("+ 1,", =-《〉-〈;〉)) + (', + ⁵ )

[0135] (c) of FIG. 5 is a graph showing the predicted value obtained in step S2.

[0136] (Step S3) The filter processing unit 11 calculates the predicted value p (n, u, v pred obtained in Step S2

) And the average pixel value of block B obtained in step SI as a removal component to be removed from the processing target image, and from the pixel value p (n, u, v) of the processing target image By subtracting the removed component, the pixel value of the filter output image! Calculate (n, u, v). The calculation formula used to calculate the pixel value p ′ (n, u, v) of the filter output image of the filter processing unit 11 is as follows.

[0137] [Equation 14] pn, u, v) = p (n, u, v)-{p _pred in, u, v)-(p ^)}

[0138] (d) of FIG. 5 is a graph showing pixel values of the filter output image obtained in step S3.

[0139] Here, the average pixel value of block B in the filter output image matches the average pixel value of block B in the processing target image, that is, the filter processing unit 11 processes the average pixel value of each block. Please note that it is something that remains unchanged before and after. For this reason, it is possible to prevent blurring and flickering of moving images due to the filter processing.

[0140] The filter processing unit 11 performs the filtering process in the vertical direction after performing the filtering process in the horizontal direction as described above. The vertical filter processing operation is a force that calculates the image data of block B with reference to block B adjacent to the lower side of block Bn.

The calculation method is the same as the horizontal filtering process, so the explanation is omitted.

[0141] (Details of inverse filtering)

Next, the inverse filter processing performed by the inverse filter processing unit 12 included in the moving image encoding device 300 and the moving image decoding device 400 will be described. In the following description, it is assumed that the inverse filter process is executed every time encoding or decoding process is performed for one screen. That is, when the inverse filter process is executed, a decoded image obtained as the sum of the difference image restored by the IDCT unit 5 and the predicted image is recorded in the frame memory 7 for one screen.

[0142] Similar to the filter processing unit 11, the inverse filter processing unit 12 executes the horizontal direction reverse filter processing and the vertical direction reverse filter processing by dividing the processing target image into a plurality of blocks. Here, the division pattern in which the inverse filter processing unit 12 divides the processing target image is the same as the division pattern in which the filter processing unit 11 divides the processing target image.

[0143] The inverse filter processing unit 12 calculates the image data of the block B from the image data of the block B in the processing target image and the image data of the block adjacent to the block B. Specifically, in horizontal filtering, the image data of block B Is calculated by referring to the block _{+ 1} image data adjacent to the right side of the block and the image data of block B in the vertical filter processing. Calculated by referring to the image data of block B adjacent below B n n + W

To do.

[0144] Hereinafter, the inverse filter processing calculation executed by the inverse filter processing unit 12 in the horizontal inverse filter processing will be described with reference to FIGS. 6 and 7. FIG. In the following, the inverse filter processing calculation for calculating the pixel value (for example, the luminance level) of one block B will be described. However, the inverse filter processing unit 12 performs all the filtering processing operations described below. Repeat for two adjacent blocks to complete horizontal filtering

Yes

FIG. 6 is a flowchart for schematically explaining the inverse filter processing calculation executed by the inverse filter processing unit 12. As shown in FIG. 6, the filtering process by the inverse filter processing unit 12 includes a step T1 for calculating an average pixel value, a step T2 for calculating a predicted value, and a step T3 for calculating an inverse filter output image. Is included.

[0146] Steps T1 to T3 will be described in more detail as follows. In the following explanation, among the 16 pixels arranged in 4 rows and 4 columns belonging to block Β, the inverse filter processing operation for calculating the pixel values of 4 pixels arranged in the V row! /, State. The inverse filter processing unit 12 performs the inverse filter processing operation described below on the first row of the block 、, the fourth row, in sequence or in parallel, so that all the pixels belonging to the block の are processed. The pixel value is calculated.

[0147] (Step T1) The inverse filter processing unit 12 calculates the average pixel value <Ρ> of the four pixels arranged in the V row of the block における in the processing target image and the V row of the block における in the processing target image. Calculate the average pixel value <Ρ> of the arranged four pixels. Inverse filter processing η + 1,

Part 12 displays the average pixel values <Ρ> and <? The following formula is used to calculate π and η + 1,

It is as follows.

[0148] [Equation 15]

Here, P (n, u, v) is a pixel value in the pixel (n, u, v) of the processing target image. (A) in FIG. 7 is a graph showing the pixel values of the image to be processed, and (b) in FIG. 7 is a graph showing the average pixel values obtained by! / In step T1 above. .

[0150] (Step T2) The inverse filter processor 12 determines that the average pixel value obtained in Step T1 <P

> And <? ) And n + 1 for each pixel in the Vth row of block Bn,

Predicted value P (n, u, v) and predicted value P (n +

pred n + 1 1 pred l, u, v) is calculated. The inverse filter processing unit 12 predicts the predicted value P (n, u, v P (n + l, u

The formula used to calculate pred pred, V) is as follows:

[0151] [Equation 16]

[0152] [Equation 17]

P ("+1,", = 1 <<--,) "+ 3 ,,> + 5 <,)

[0153] (c) of FIG. 7 is a graph showing the predicted value obtained in step T2.

[0154] (Step T3)

The inverse filter processing unit 12 calculates the predicted value P (n, u, v) obtained in step T2 and the step pred

The difference from the average pixel value of block B obtained in T1 is set as an additional component to be added to the processing target image, and the additional component is added to the pixel value P (n, u, v) of the processing target image. By doing so, the pixel value P ′ (n, u, V) of the filter output image is calculated. The calculation formula used by the inverse filter processing unit 12 to calculate the pixel value P ′ (n, u, v) of the inverse filter output image is as follows.

[0155] [Equation 18]

[0156] (d) of FIG. 7 is a graph showing pixel values of the inverse filter output image obtained in step T3. Note that the average pixel value of block B in the inverse filter output image matches the average pixel value of block B in the processing target image. This Therefore, occurrence of blurring and flickering of moving images is effectively prevented.

[0157] The inverse filter processing unit 12 performs the inverse filter processing in the horizontal direction as described above, and then performs the inverse filter processing in the vertical direction. The vertical inverse filtering calculation is to calculate the image data of block B with reference to block B adjacent to the lower side of block B n n + W

Since the calculation method is the same as that of the horizontal inverse filter processing, the description thereof is omitted.

The inverse filter processing performed by the inverse filter processing unit 12 defined as described above corresponds to the inverse transformation of the filter processing performed by the filter processing unit 11. That is, p ′ (n, u, v) = P (n between the output p ′ (n, u, v) of the filter processing unit 11 and the input P (n, u, v) of the inverse filter processing unit 12 , U, v), the relationship between the input p (n, u, v) of the filter processor 11 and the output P '(n, u, v) of the inverse filter processor 12 is p (n , U, ν) = Ρ '(η, u, ν). Accordingly, the frequency component that generates the block distortion removed from the processing target image by the filter processing of the filter processing unit 11 is restored by the inverse filter processing of the inverse filter processing unit 12. Therefore, a specific frequency component is not lost in the restored image.

Yes

[0159] In addition, when using a prediction in which the average pixel value of the prediction value for each block is different from the average pixel value of the input image for each block, the average pixel value of the prediction value and the average pixel value of the input image are used as the filter output. It is preferable to make a correction so that the average pixel value of the filter input and filter output is maintained.

[Problems in the basic configuration]

As described above, the moving picture coding apparatus 300 removes the specific frequency component FL that causes block distortion by performing the filtering process on the coding target image by the filter processing unit 11. Then, the video decoding device 400 restores the removed frequency component FL by the inverse filter processing unit 12. Thereby, block distortion in the inverse filter output image in the moving picture decoding apparatus 400 is reduced.

[0161] At this time, in order for the inverse filter output image output from the inverse filter processing unit 12 to match the encoding target image input to the filter processing unit 11, the decoding input to the inverse filter processing unit 12 is performed. The image must match the filter output image output from the filter processor 11 There is. In other words, in order to accurately restore the frequency component FL removed in the inverse filter output image, the filter output image and the decoded image need to match exactly. Speaking of Fig. 5 and Fig. 7, the graph shown as (d) in Fig. 5 should match the graph shown as (a) in Fig. 7.

However, the filter output image in moving picture coding apparatus 300 and the decoded picture in moving picture decoding apparatus 400 are different from each other by a quantization error. This is achieved by quantization in the video encoding device 300 by the DCT unit 1 and the quantization unit 2 and inverse quantization in the video decoding device 400 by the inverse quantization unit 4 and the IDCT unit 5. This is because a quantization error occurs between the quantization target image and the restored quantization target image. For this reason, the restoration of the frequency component FL by the inverse filter processing unit 12 is approximate. Further, the same can be said for the inverse filter processing performed for generating the reference image inside the moving image coding apparatus 300.

[0163] (a) to (d) of FIG. 17 are respectively input to the encoding target image input to the filter processing unit 11, the filter output image output from the filter processing unit 11, and the inverse filter processing unit 12. 5 is a graph showing the pixel values of the decoded image and the inverse filter output image output from the inverse filter processing unit 12. As can be seen by comparing the graph shown in FIG. 17 (b) with the graph shown in FIG. 17 (c), the pixel value of the decoded image has a quantization error relative to the pixel value of the filter output image. It has. FIG. 17D is a graph of the inverse filter output image in FIG. 17D, which is the result of applying the inverse filter process to the decoded image including the quantization error. As shown in (d) of FIG. 17, the inverse filter output image does not accurately reproduce the encoding target image. That is, the frequency component reconstructed by the inverse finisher processing unit 12 has an error with respect to the frequency component FL removed by the filter processing unit 11.

[Embodiment 1]

An embodiment of the present invention will be described below with reference to FIGS. 1 and 8 to 10.

[0165] (Configuration of video encoding device)

FIG. 8 is a functional block diagram showing a schematic configuration of the moving picture coding apparatus 300a according to the present embodiment. As shown in FIG. 8, the moving picture coding apparatus 300a includes a DCT unit 1 and a quantization unit. 2, variable length encoding unit 3, inverse quantization unit 4, IDCT unit 5, frame memory 7, prediction image derivation unit 8, encoding control unit 21, filter processing unit 11, inverse filter processing unit 12, correction determination unit 13 A corrected image candidate derivation unit 14 and a corrected image determination unit 15 are provided. In FIG. 8, blocks having the same functions as those in the moving picture coding apparatus in FIG. 2 are denoted by the same names and symbols as in FIG.

[0166] A feature of the moving image coding apparatus 300a is that it includes a correction determination unit 13, a corrected image candidate derivation unit 14, and a corrected image determination unit 15.

[0167] The corrected image determination unit 15 performs correction for subtracting the corrected image from the quantization target image for at least one block. As shown in FIG. 8, the corrected image determining unit 15 may subtract the corrected image from the quantization target image by adding the corrected image to the predicted image.

[0168] The correction determination unit 13 determines a block on which the correction image determination unit 15 should perform the correction. Specifically, a block in which the average pixel value of the quantization target image after the correction is performed is smaller than the average pixel value of the quantization target image before the correction is performed should be corrected. Judge as a block. The corrected image determination unit 15 performs the above correction on the blocks that are determined to be corrected by the correction determination unit 13.

[0169] The corrected image candidate derivation unit 14 derives correction image candidates used by the corrected image determination unit 15 for the correction. More specifically, a corrected image candidate having an approximate average pixel value that approximates the average pixel value of the quantization target image in each block as a pixel value is derived as a corrected image candidate.

[0170] For this reason, the corrected image determination unit 15 makes the average pixel value of the quantization target image after the correction is smaller than the average pixel value of the quantization target image before the correction is performed. For the block, correction is performed by subtracting the approximate average pixel value that approximates the average pixel value of the quantization target image in the block to which the pixel belongs from the pixel value of each pixel of the quantization target image.

[0171] Therefore, according to the correction described above, the average pixel size of each block of the quantization target image can be reduced to be equal to or less than the original average pixel value of the block. Therefore, the quantization level interval in the above quantization can be set small, and is generated by quantization. Quantization error can be reduced. For this reason, it is possible to reduce the quantization error generated in the quantization process. Therefore, the quantization target image after the filtering process and the restored quantization target image before the inverse filtering process can be matched more accurately. For this reason, the frequency component removed by the filtering process can be restored more accurately during the inverse quantization.

[0172] (Configuration of video decoding device)

FIG. 1 is a functional block diagram illustrating a schematic configuration of a video decoding device 400a corresponding to the video encoding device 300a shown in FIG. As shown in FIG. 8, the video decoding device 400a includes an inverse quantization unit 4, an IDCT unit 5, a frame memory 7, a predicted image derivation unit 8, an inverse filter processing unit 12, a variable length decoding unit 20a, and a corrected image candidate. A derivation unit 14 and a corrected image determination unit 15 are provided.

[0173] The difference between the moving picture decoding apparatus 400a and the moving picture decoding apparatus 400 (Fig. 3) is that the moving picture decoding apparatus 400a includes a corrected image candidate deriving unit 14 and a corrected image determining unit 15. is there. Further, in addition to the function of the variable-length decoding unit 20, the variable-length decoding unit 2 Oa provided instead of the variable-length decoding unit 20 of the video decoding device 400 provides information on whether or not to correct the predicted image. That is, the correction determination result is decoded and supplied to the corrected image candidate deriving unit 14 and the corrected image determining unit 15. Note that the corrected image candidate derivation unit 14 and the corrected image determination unit 15 in the moving image decoding device 400a have the same functions as those indicated by attaching the same reference numerals to the moving image encoding device 300a, respectively. .

[0174] The moving picture decoding apparatus 400a can decode the encoded data encoded by the moving picture encoding apparatus 300a. Since the above-described correction is used at the time of encoding in the moving image encoding device 300a, the quantization error in the input image to the inverse filter in the moving image decoding device 400a is reduced. Therefore, since the removed frequency component FL can be accurately restored by the inverse filter processing, the block distortion in the inverse filter output image of the inverse filter processing, which is also the output image of the video decoding device 400a, is reduced. That's the power S.

Hereinafter, the correction determination process at the time of encoding block B will be described. The corrected image derivation process at the time of encoding and decoding of block B will be described. In the following description, the encoding process of the input image or the decoding process of the encoded data (in FIG. 18, the lock, lock B, pre, lock B, pre, lock B, ... , Lock B, pu, lock B

0 1 2 n n

, ... and! /, Which are executed in raster scan order.

+ 1

[0176] (Details of correction judgment processing)

The correction determination process is executed in the correction determination unit 13. First, the correction determination unit 13 that can be suitably used for the moving image encoding device 30 Oa will be described.

The correction determination unit 13 includes a filter output image output from the filter processing unit 11, a predicted image output from the predicted image derivation unit 8, and a corrected image candidate output from the correction image candidate derivation unit 14. And are supplied. Based on these, the correction determination unit 13 executes a correction determination process described later. The determination result is sent as a correction determination result to the corrected image candidate deriving unit 14, the corrected image determining unit 15, and the variable length encoding unit 3. The variable length encoding unit 3 encodes the correction determination result and sends it to the moving image encoding apparatus 400a.

FIG. 9 is a functional block diagram showing a schematic configuration of the correction determination unit 13. As shown in FIG. 9, the correction determination unit 13 includes an average pixel value calculation unit 50 and a corrected image candidate evaluation unit 51.

[0179] In the average pixel value calculation unit 50, a difference image between each pixel of the encoding target image (filter output image) subjected to the filter process and the prediction image supplied from the prediction image derivation unit 8 is stored. Entered. The average pixel value calculation unit 50 calculates an average pixel value for each block of the input difference image. Here, the difference image corresponds to the quantization target image when the correction is not performed on the currently focused block. Therefore, the calculated average pixel value is obtained when the correction is not performed. Note that this corresponds to the average pixel value of the quantization target image. The average pixel value of the difference image calculated by the average pixel value calculation unit 50 is sent to the corrected image candidate evaluation unit 51.

[0180] The corrected image candidate evaluation unit 51 subtracts the average pixel value of the difference image supplied from the average pixel value calculation unit 50 and the correction image candidate supplied from the correction image candidate derivation unit 14 from the difference image. The average pixel value of the obtained image is compared, and it is determined whether or not correction is to be performed in the block. Specifically, when the average pixel value of the image obtained by subtracting the corrected image candidate from the difference image is smaller, correction is performed on the block B currently being processed. Judge that. Here, it should be noted that the image obtained by subtracting the corrected image candidate supplied from the corrected image candidate deriving unit 14 from the difference image corresponds to the quantization target image when correction is performed. Therefore, the corrected image candidate evaluation unit 51 determines that a block that can reduce the average pixel value of the quantization target image as a block to be corrected by performing the correction. The determination result in the corrected image candidate evaluation unit 51 is supplied to the corrected image determination unit 15 and the variable length coding unit 3 as a correction determination result.

Next, the procedure of the correction determination process executed in the correction determination unit 13 will be described. The procedure of the correction determination process includes two steps: a step of calculating an average pixel value of a difference image that is a difference between the encoding target image and the predicted image, and a step of determining whether or not to perform correction.

[0182] (Step 1) The average pixel value calculation unit 50 calculates the pixel value P (n, u) of the predicted image from the pixel value p (n, u, v) of the encoding target image after the filtering process. , v) subtraction image est

An element is acquired, and an average pixel value of the difference image is calculated. The calculation formula used to calculate the average value in block B is as follows.

sub, n

[0183] [Equation 19]

<Ps «l ,, n ^> ― Pes, ^U ')

[0184] (Step 2) The corrected image candidate evaluation unit 51 adds sub to the average pixel value calculated in Step 1 and the average pixel value of the corrected image candidate supplied from the corrected image candidate derivation unit 14. n

A comparison is made with the corresponding correction amount a ′ to determine whether or not the correction is performed in the block B. Specifically, the determination is made by comparing a and 0 represented by the following mathematical formula.

[0185] [Equation 20]

[0186] [Equation 21] β = \ <Ρ „> \

[0187] If α is smaller than / 3, it is determined to perform correction. In other cases (when α is greater than / 3, or when α and / 3 are equal), it is determined that correction is not performed. As described above, α is an amount corresponding to the average pixel value of the quantization target image when correction is performed, and β is an amount corresponding to the average pixel value when correction is not performed. Therefore, by comparing the magnitudes of α and / 3, it is possible to determine whether or not the average pixel value of the quantization target image can be reduced when correction is performed.

[0189] (Details of corrected image derivation process)

Next, a corrected image derivation process executed in the corrected image candidate derivation unit 14 and the corrected image determination unit 15 will be described. First, the corrected image candidate deriving unit 14 and the corrected image determining unit 15 that can be suitably used in the moving image encoding device 300a and the moving image decoding device 400a will be described.

[0190] The corrected image candidate derivation unit 14 is based on the correction determination result and the conversion coefficient supplied from the inverse quantization unit 4, and is a corrected image candidate that is a correction image candidate used for correcting the block. Is derived. The derived corrected image candidates are sent to the correction determination unit 13 and the corrected image determination unit 15.

[0191] Based on the correction determination result supplied from the correction determination unit 13, the correction image determination unit 15 determines and transmits a correction image in the block.

FIG. 10 is a functional block diagram showing a schematic configuration of the corrected image candidate derivation unit 14. Figure 1

As indicated by 0, the corrected image candidate derivation unit 14 includes a prediction error amount storage memory 52, a correction amount derivation unit 53, a corrected image candidate generation unit 54, and a prediction error amount derivation unit 55.

[0193] The prediction error amount storage memory 52 stores the prediction error amount corresponding to the quantized / inverse quantized block (for example, block B) processed before the block B currently being processed. .

The correction amount deriving unit 53 refers to the prediction error amount recorded in the prediction error amount storage memory 52 and derives the correction amount in the block. The derived correction amount is sent to the corrected image candidate generation unit 54.

The corrected image candidate generating unit 54 generates and sends a corrected image candidate based on the correction amount supplied from the correction amount deriving unit 53.

[0196] The prediction error amount deriving unit 55 is based on the correction determination result supplied from the correction determining unit 13, the correction amount supplied from the correction amount deriving unit 53, and the transform coefficient supplied from the inverse quantization unit 4. !! The prediction error amount in the block currently being processed is calculated and recorded in the prediction error amount storage memory 52.

Next, the corrected image derivation process executed by the corrected image candidate deriving unit 14 and the corrected image determining unit 15 will be described with reference to FIG.

FIG. 11 is a flowchart showing the flow of the corrected image derivation process.

[0199] The corrected image derivation process includes a step U 1 for deriving a correction amount applicable to the block B, a step U 2 for generating a corrected image candidate based on the derived correction amount value, and a correction determination. Step U3 for making a determination based on the result, Step U4 for sending a corrected image candidate as a correction amount, and Step U5 for sending an image with all pixel values being zero as a correction amount. Hereinafter, steps U1 to U5 will be described in a little more detail.

[0200] (Step U1) The correction amount deriving unit 53 sets the prediction error amount d corresponding to the block B− adjacent to the left side of the block B currently being processed as the correction amount of the block B. However, if the block B is located at the left edge of the image, the value zero is set as the correction amount for the block.

[0201] [Equation 22] d \ also —, (("mod W) ≠ 0)

r ^t — 1 θ ((«mod W) = 0)

[0202] Here, the reason for setting the prediction error amount d of block B to the correction amount of the block B is that the prediction image is corrected by using the correlation of the prediction error amount between adjacent blocks. This is because the accuracy of the image can be increased.

[0203] (Step U2) The corrected image candidate generation unit 54 generates a corrected image candidate based on the correction amount a ′ derived in Step U1. Specifically, an image in which the pixel values of all the pixels are a ′ is generated, and the image is set as a corrected image candidate.

[0204] (Step U3) The corrected image determination unit 15 is based on the correction determination result supplied from the outside (the correction determination unit 13 in FIG. 8 or the variable length decoding unit 20a in FIG. 1)! /, Determine the corrected image

. If the determination result indicates that correction is performed in the block, the process proceeds to step U4.

If no correction is made! /, Go to step U5.

[0205] The correction determination result used for the determination in this step is the same as that in the encoding process. Is supplied from the correction determination unit 13 and is supplied from the variable length decoding unit 20a during the decoding process.

(Step U4) This step is executed when it is determined that correction is performed on the block currently being processed as a result of the prediction determination in step U3. The corrected image candidate generation unit 54 sends out the corrected image candidate derived in step U2 as a corrected image of the block.

(Step U5) This step is executed when it is determined that correction is not performed as a result of the prediction determination in step U3. The corrected image candidate generation unit 54 generates an image in which the pixel values of all the pixels are zero, and sends the image as a corrected image candidate.

[0208] According to the above procedure, a corrected image candidate necessary for performing correction in the block currently being processed can be derived. In the moving image encoding device 300a and the moving image decoding device 400a, correction is performed by adding the derived corrected image to the predicted image transmitted from the predicted image deriving unit 8.

[0209] In the above-described corrected moving image derivation process, the prediction error amount d of block B- is referred to in deriving the correction amount of block B currently being processed. Therefore, the prediction error amount d of the block B may also be used for derivation of a correction amount in a block to be processed later (for example, block B). Therefore, the prediction error amount d is derived and recorded in the prediction error amount recording memory 52. It is necessary to keep it. The procedure for deriving the prediction error amount d is as follows.

[0210] (Step VI) The prediction error amount deriving unit 55 includes the correction amount value derived in step U1 of the corrected image deriving process, the correction determination result supplied from the correction determination unit 13, and the inverse quantization unit 4 The prediction error amount d is calculated on the basis of the transform coefficient after inverse quantization supplied from the above. The calculation formula used for calculation is as follows.

[021 1] [Equation 23] r (When prediction correction is not performed in the block S „)

_r + a „'(when the prediction is corrected in the block„)

Here, r is a DC component in the transform coefficient after inverse quantization, and r X γ corresponds to the average pixel of the quantization target image in block Β. Note that _Ί is a constant for converting the unit of the conversion coefficient r after inverse quantization to the unit of the pixel value. Also quantity Since the correction performed on the child image is to subtract the correction amount α _η 'from each pixel of the image to be quantized, the quantization before correction is performed until r X y + a This corresponds to the average pixel value of the target image (difference image obtained by subtracting the predicted image from the target image to be encoded). That is, the prediction error amount d corresponds to the average pixel value of the quantization target image regardless of whether correction is performed in block B or correction is performed.

[0213] (Step V2) The prediction error amount deriving unit 55 records the prediction error amount d derived in step VI in the prediction error amount recording memory 52. At this time, a prediction error amount that is not used in a block processed after the block may be deleted from the prediction error amount storage memory 52.

[0214] (Additional note regarding corrected image derivation processing)

1. Describes a configuration in which correction determination results for all blocks are encoded on the encoding side, and correction determination results for all blocks are obtained by decoding the encoded correction determination results on the decoding side. However, the present invention is not limited to this.

[0215] That is, the correction determination result for all blocks is not necessarily required on the decoding side. For blocks whose correction determination result can be estimated on the decoding side, encoding of the correction determination result is omitted. A configuration is also possible. By omitting the encoding of the correction determination result for the block on which the correction determination result can be estimated on the decoding side, the code amount of the encoded data can be reduced.

[0216] For example, correction is not always performed for a block in which the derived correction amount compensation value is zero. Therefore, it is possible to estimate (determine) the correction determination result on the decoding side by referring to the correction amount. Also, for example, correction tends to be difficult to apply to blocks that have a certain amount of correction value. Therefore, when the value of the correction amount is larger than a predetermined threshold value, it can be estimated that correction is not performed. Therefore, the encoding side can be configured to omit encoding for such blocks.

[0217] 2. Correction amount deriving section 53 A power that explains the configuration that uses the prediction error amount d corresponding to block B- adjacent to the left side of block B to derive the correction amount. Akira is not limited to this. Specifically, for example, a prediction error amount corresponding to another processed block existing in the vicinity of the block may be used together.

For example, when encoding and decoding processes are performed in the raster scan order as shown in FIG. 20, the prediction error amount d corresponding to the block B adjacent above the block B and the block B The correction amount can be derived by using the prediction error d corresponding to block B- adjacent on the left side. In this case, by selecting a block having a prediction error amount that is highly correlated with the prediction error amount of the block B from two adjacent blocks, the prediction error amount can be reduced by one. As a result, more appropriate correction can be made. Hereinafter, this example will be described in more detail.

[0219] In order to use the prediction error amount of the block on the left side and the upper side of the block for derivation of the correction amount, step U1 in the correction image derivation process described above may be replaced with the next step ur.

[0220] (Step UD correction amount deriving unit 53 determines, based on prediction error amount d corresponding to block Β adjacent to the left side of block Β and prediction error amount d corresponding to block B adjacent to the upper side, Calculate the correction amount, which is determined by the following equation.

[0221] [Equation 24]

0 {ά _η _ _χ = 0 or << mod W = 0) and (d. One ^ = 0 or w <) (d „one, ≠ 0 and n mod W ≠ 0) and (d„ _—w = 0 or n W W) (d „_ _w ≠ 0 and w> if) and (—! = 0 or n mod W = -0) —, + d„

≠ 0 and n mod W ≠ 0) and (0 and n> W

2

[0222] Note that the prediction errors corresponding to the two blocks B and B existing on the left side of the block B and the two blocks B and B existing on the upper side of the block B, respectively.

-2W

Based on d, the correction amount a 'can be calculated using the following formula: n-2W n

[0224] However, in the case of n <2W or n mod W <2, it is obtained according to the above [Equation 24]. [0225] By calculating the correction amount as described above, it is determined whether the correlation of the prediction error amount between the blocks is high in the vertical direction or the horizontal direction of the block, and the correlation is high with respect to the block The correction amount can be derived using the prediction error amount of the block located in the direction. For this reason, more accurate correction can be performed.

[0226] In the above description, the correction amount is derived based only on the prediction error amount. However, the correction amount may be derived using a coding parameter correlated with the prediction error amount. . For example, it is possible to derive the correction amount by combining the prediction mode used for intra prediction and the prediction error amount. More specifically, a block whose prediction mode matches the prediction mode in the block is selected from blocks adjacent to the block, and the correction amount is calculated as an average value of the prediction error amount in the selected block. The ability to seek S Predictive images are often similar in blocks with the same prediction mode, so the amount of prediction error is often near and low. Therefore, a more preferable correction amount can be obtained by using the prediction error amount of the block having the same prediction mode.

[0227] Coding parameters correlated with the prediction error amount include, in addition to the prediction mode, a block size for intra prediction, a motion vector in inter prediction, a reference image, and the like. For example, a preferable correction amount can be obtained by selecting a block having the encoding parameter closest to the block from the neighboring blocks of the block and obtaining the value of the correction amount a ′ from the prediction error amount of the selected block. Can be sought.

[Embodiment 2]

The following describes the second embodiment of the present invention with reference to FIGS.

[0229] (Configuration of video encoding device)

FIG. 12 is a functional block diagram showing a schematic configuration of the moving picture coding apparatus 300b according to the present embodiment.

[0230] As shown in Fig. 12, the moving picture coding apparatus 300b includes a DCT unit 1, a quantization unit 2, a variable length coding unit 3, an inverse quantization unit 4, an IDCT unit 5, a frame memory 7, and a prediction Image deriving unit 8, coding control unit 21, filter processing unit 1 1, inverse filter processing unit 12, inverse filter input correction unit 16, And a correction method determination unit 17.

[0231] The difference between the moving picture coding apparatus 300b and the moving picture coding apparatus 300 (Fig. 1) is that the moving picture coding apparatus 300b uses an inverse filter that corrects a local decoded image to be subjected to inverse filter processing. This is the point that a correction unit 16 is provided. In addition, the moving picture coding apparatus 300b includes a correction method determination unit 17 that determines a correction method to be applied in the inverse filter input correction unit 16.

[0232] The inverse filter input correction unit 16 acquires the local decoded image recorded in the frame memory 7, and corrects the local decoded image according to the correction method supplied from the correction method determination unit 17. The corrected local decoded image is sent to the inverse filter processing unit 12.

[0233] The correction method determination unit 17 is applied by the inverse filter input correction unit 16 based on the input image data that is an input image to the video encoding device 300a and the local decoded image acquired from the frame memory 7. Determine how to fix. The determined correction method is sent to the inverse filter input correction unit 16 and the variable length coding unit 3 as a correction method determination result.

[0234] As already described, in moving picture coding apparatus 300 and moving picture decoding apparatus 400 having the above-described basic configuration, when a locally decoded picture includes a quantization error, the frequency is reduced by inverse filtering. There was a problem that component FL could not be restored correctly!

[0235] In this embodiment, the local decoded image (Fig. 15) has a reduced quantization error by correcting the local decoded image (Fig. 15 (b)) including the quantization error. (c) Find). When inverse filtering is applied to a locally decoded image that includes quantization errors, the frequency component FL that has been removed! / Cannot be accurately restored, so the image obtained as a result of inverse filtering contains block distortion (see Fig. 2). 15 (d)). On the other hand, when the inverse filter process is applied to a locally decoded image with a reduced quantization error, the frequency component FL can be restored more accurately, so that the block distortion is reduced in the image obtained as a result of the inverse filter process. / !, ru (Fig. 15 (e)).

[0236] In the video encoding device 300b, the correction method determination unit 17 determines the power that can reduce the quantization error in the locally decoded image by a predetermined correction method, and encodes the determination result. . At the time of decoding, the quantization error of the local decoded image can be reduced by correcting the local decoded image according to the determination result. Therefore, the video encoding device 300b can generate encoded data that can be used to decode an image with reduced block distortion. [0237] (Configuration of video decoding device)

FIG. 13 is a functional block diagram showing a schematic configuration of a video decoding device 400b corresponding to the video encoding device 300b shown in FIG. As shown in FIG. 13, the video decoding device 400b includes an inverse quantization unit 4, an IDCT unit 5, a frame memory 7, a predicted image derivation unit 8, an inverse filter processing unit 12, an inverse filter input correction unit 16, and A variable length decoding unit 20b is provided.

[0238] The difference between the moving image decoding apparatus 400b and the moving image decoding apparatus 400 (Fig. 3) is that the moving image decoding apparatus 400b has a function of decoding the correction method determination result applied by the inverse filter input correction unit. The variable length decoding unit 20b is provided. Further, the moving picture decoding apparatus 400b includes an inverse filter input correction unit 16 that reduces a prediction error by correcting a local decoded picture based on the correction method determination result.

[0239] Since the moving picture decoding apparatus 400b includes the inverse filter input correction unit 16, it is possible to reduce the quantization error occurring in the local decoded picture. Therefore, in the inverse filter process, the specific frequency component FL that causes the block distortion removed by the filter process at the time of encoding can be reproduced more accurately than the moving picture decoding apparatus 400b. An image with reduced block distortion can be decoded.

[0240] Hereinafter, the correction method determination process, which is a characteristic process in the video encoding device 300b, will be described in detail. Further, the inverse filter input correction process, which is a characteristic process in the moving picture coding apparatus 300b and the moving picture decoding apparatus 400b, will be described in detail.

[0241] In the following description, for the sake of simplicity, correction method determination processing and inverse filter input are performed with a local decoded image equivalent to one screen of input image data recorded in the frame memory 7. It is assumed that the correction process is activated.

[0242] (Details of correction method decision processing)

With reference to the flowchart of FIG. 14, the procedure of the correction method determination process executed in the correction method determination unit 16 of the video encoding device 300b will be described. Steps W;! To W6 represent processing for determining a correction method for the block of interest B in the locally decoded image.

Yes [0243] (Step Wl) The correction method determination unit 17 calculates the average pixel value P> of the encoding target image in the block of interest Bn. FIG. 15 (a) is a diagram schematically showing the average pixel value P> of the encoding target image obtained by this step.

(Step W2) A locally decoded image is acquired from the frame memory 7, and an average pixel value P> of the locally decoded image in the target block B is calculated. (B) of Fig. 15 is a diagram schematically representing the average pixel value P> of the locally decoded image obtained in step W2.

(Step W3) In order to correct the pixel value of the locally decoded image in the target block B, a corrected value is calculated from the pixel value of the locally decoded image. First, the left of the attention block

mod, n

Find the average pixel values <P-> and of the local decoded image in the two blocks adjacent to the right. Then, calculate the weighted average of the average pixel values P −> and of the obtained local decoded image and the average pixel value of the local decoded image in the target block obtained in step W2. Thus, the corrected value is obtained. More specifically,

πιοα, η

Calculate the corrected value <Ρ> using the following formula.

πιοα, η

[0246] [Equation 26]

〈 _D , „〉 = ^ (2,〉 + 〈+ 〈)

[0247] When the target block B is positioned at the screen edge, the following calculation method is used.

[0248] [Equation 27]

〈,,,〉 = (〈„〉 + 〈) (» Modf = 0)

[0249] [Equation 28],〉 = (,〉 ten 〈) (nmodW = W-l)

[0250] Note that the above calculation is performed by smoothing the average pixel value of the decoded image in the block of interest and its neighboring blocks, so that local decoding in the block of interest when no quantization error occurs is performed. Calculate the corrected value that approximates the average pixel value of the image

mod, n

It corresponds to the processing. The above correction is particularly suitable for reducing the quantization error generated in an image region where the fluctuation of the pixel value is flat. Figure 15 (c) is based on step W3. The average pixel value of the modified local decoded image obtained

lnod, n

is there.

[0251] (Step W4) The average pixel value of the encoding target image, the average pixel value of the local decoded image, and the step W3, respectively, derived in Step W1 and Step W2, respectively. Based on the corrected value <Ρ '>, it is determined whether or not to correct the pixel value in the target block Β. If the value of the modified value <値 ′> is closer to the value of <ρ> of the encoding target image than the average pixel value <Ρ> of the local decoded image, the process proceeds to step W5. Otherwise go to step W6.

[0252] (Step W5) It is determined that the target block Β is to be corrected, and this determination result is sent as a correction method determination result to the inverse filter input correction unit 16 and the image encoding unit 3 to be corrected. The decision process is terminated.

[0253] (Step W6), the target block 修正 is determined not to be corrected! /, And the determination result is sent to the inverse filter input correction unit 16 and the image encoding unit 3 as a correction method determination result. Then, the correction method determination process ends.

[0254] By performing the correction method determination process according to the above procedure, the correction method to be applied to the block 入力 by the inverse filter input correction unit 16 is determined, and the correction method is transmitted as the correction method determination result. The power to do S. In the inverse filter input correction process described later, correction can be made only when the quantization error of the block of interest can be reduced according to the determination result.

[Details of inverse filter input correction processing]

Next, an inverse filter input process executed in the inverse filter input correction unit of the moving image encoding device 300b and the moving image decoding device 400b will be described based on the flowchart of FIG. Steps X;! To X7 represent processing for calculating the pixel value of the target block Bn in the input image to the inverse filter.

(Step XI) A locally decoded image is acquired from the frame memory 7.

[0257] (Step X2) The correction method determination result derived in the correction method determination process and supplied from the correction method determination unit 17 or the variable length decoding unit 20b is the! If you indicate that you want to make corrections, go to step X3. The judgment result is Proceed to Step X7 to indicate that the local block image is to be corrected for eye block Bn! /.

(Step X3) The average pixel value of the locally decoded image in the target block Bn obtained in Step XI is calculated.

[0259] (Step X4) The correction value <

Calculate the value of P>.

niod, n

[0260] (Step X5) Using the average pixel value P> obtained in Step X3 and the modified value obtained in Step X4, the pixel value P ( n, u, V)

niod, n mod

Is calculated by the following calculation formula.

[0261] [Numerical 29]

P _mai (w) = P (n, u, v) + P <> [0262] where 1 is the quantum from the average pixel value of the local decoded image. Conversion error

n mod, n

This indicates the difference obtained by subtracting the correction value that approximates the average pixel value of the local decoded image in the target block when it does not occur, that is, the approximate quantization error amount in block B. Therefore, the correction corresponds to a process of removing the approximate quantization error from the pixel value of each pixel of the locally decoded image. With this modification, the average pixel value Pn of the locally decoded image in block B is replaced with the modified average pixel value niod, n obtained in step X2.

[0263] (Step X6) The corrected local decoded image obtained in Step X5 is sent to the inverse filter processing unit 12, and the inverse filter input correction process is terminated.

(Step X7) The local decoded image acquired in Step XI is sent as it is to the inverse filter processing unit 12, and the inverse filter input correction process is terminated.

[0265] By performing the inverse filter input correction process according to the above procedure, the locally decoded image can be corrected and the quantization error can be reduced. By applying the inverse filter process to the locally decoded image that has been modified and the quantization error is reduced, it is corrected and removed by filtering as compared to the case of inputting the locally decoded image. The force S can be restored more accurately to restore the frequency component. [Additional Notes about Embodiment 2]

In the above description regarding the correction method determination processing in the moving image encoding device 300b, the correction method determination unit 17 is assumed to send the correction method determination result to the variable length encoding unit 3. However, it is not always necessary to encode the correction method determination results for all the blocks. If the correction method determination results can be estimated at the time of decoding, the encoding may be omitted.

[0267] For example, in the correction method determination process, when the average pixel value <Pn> of the local decoded image in block Bn matches the average pixel value of the corrected local decoded image

niod, n

Therefore, it is always determined that the inverse filter input correction unit 16 does not perform correction. In the case described above, since the determination result can be easily estimated at the time of decoding, it is not necessary to encode the determination result! /. Therefore, the code amount of encoded data can be reduced.

[0268] If there is a difference between the average pixel value of the locally decoded image and the average pixel value of the modified local decoded image to some extent, the average pixel value is better. Input

niod, n n

There is a tendency that the average pixel value of image data is close to p>. Therefore, a certain threshold T1 is determined, and and <? > When ST1 is greater than ST1, the local decoded image is always estimated.

n mod, n

It is determined not to perform. In that case, and <? Evaluating the difference between

n mod, n

Therefore, it is not necessary to encode the correction method determination result. As the threshold T1, for example, a value represented by the following formula can be used.

[0269] [Equation 30]

[0270] where r is the quantization used to quantize the DC component value in the DCT transform coefficient

0

The step width, γ, is used to convert the value of r from frequency domain units to spatial domain units.

0 0

It is a coefficient. The threshold T1 is determined as described above due to the fact that the difference between the average pixel value P> and the average pixel value P> of the input image data is basically less than r X γ.

η 0 0

[0271] In order to derive the average pixel value <Ρ> of the modified local decoded image,

niod, n

When smoothing with reference to the value of the average pixel value is used, there is a tendency that correction is easily performed on an image area where the fluctuation of the pixel value is flat. Therefore, for example, quantization representative value To determine whether or not the target block is a flat region. If it is determined that the target block is a flat region, the correction method determination result is not encoded and the correction method determination result is not encoded. The amount can be reduced.

[0272] In the correction method determination process and the inverse filter input correction process described above, the left and right sides of the target block are used to derive the average pixel value in the corrected local decoded image.

niod, n

Force using smoothing with blocks adjacent to the average pixel value

mod, n may be derived.

[0273] For example, the average pixel value can be derived by smoothing using four blocks adjacent to each of the top, bottom, left, and right of the block of interest. Average image of the block of interest Bn

niod, n

Using the prime value and the average pixel values <P->, , , of blocks adjacent to the top, bottom, left, and right of block Bn, the corrected average pixel value P>

The value of W n + W n− 1 n + 1 mod, n can be calculated by the following calculation formula.

[0274] [Equation 31]

〈> = (4 〈„〉 + 〈+〉 + 〈!〉 + 〈)

[0275] When the target block B is located at the screen edge, the calculation is performed by another calculation method. The calculation formula when block B is at the upper end (except for the upper left corner and upper right corner) and the upper left corner is shown below.

[0276] [Equation 32]

〈 _D , „〉 =〉 +〉 + 〈"-+ 〈"(" Katsu and "≠. And-1)

[0277] [Equation 33],〉 =-(2 („) + („ _{+ f} ) + („ ₊₁ ) j (" = 0)

[0278] The case where block B is located at a screen edge other than the above can be calculated based on the same concept, and the description is omitted.

[0279] [Appendix]

The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope shown in the claims, and the technical means disclosed in the different embodiments are applied appropriately. Embodiments obtained by appropriate combinations are also included in the technical scope of the present invention. For example, the present invention can be configured as follows.

[0280] The video decoding device according to the present invention is a video decoding device that decodes encoded data obtained by encoding image data from which frequency components that generate block distortion have been removed.

Variable length decoding means for decoding the encoded data;

An inverse filter processing means for performing an inverse filter process for restoring the removed frequency component on the image data obtained by decoding;

Prediction image generation means for generating a prediction image based on pixel values of a reference image; prediction error amount derivation means for deriving a prediction error amount approximating the difference between the image data from which the frequency component has been removed and the prediction image; ,

Correction means for correcting the predicted image based on the prediction error amount;

It is configured with! /

[0281] Further, in the video decoding device according to the present invention,

The variable length decoding means decodes the correction judgment result indicating whether or not to correct the correction means!

The correction means switches whether to correct the predicted image based on the prediction error amount according to the correction determination result.

It may be configured as follows.

[0282] Also, in the video decoding device according to the present invention,

The correction means estimates whether or not the predicted image can be corrected based on the prediction error amount! /, Based on the prediction error amount in the processing target block, and when the estimation is possible, Depending on the result of the estimation, based on the prediction error amount! /, It is switched whether to correct the predicted image.

It may be configured as follows.

[0283] Also, in the video decoding device according to the present invention,

The correction means corrects the prediction image based on the prediction error amount corresponding to at least one decoded block! It may be configured as follows.

[0284] Also, in the video decoding device according to the present invention,

The decoded block includes at least one block adjacent to the processing target block,

It may be configured as follows.

[0285] Also, in the video decoding device according to the present invention,

The correction means corrects the prediction image based on the prediction error amount and an encoding parameter correlated with the prediction error amount.

It may be configured as follows.

[0286] The moving image decoding apparatus according to the present invention divides an image into a plurality of blocks, quantizes the image data of each block, and encodes a quantized representative value obtained by the quantization. Encoding device,

Filter processing means for performing filter processing for removing frequency components that generate block distortion on the image data before quantization;

A predicted image generating unit that generates a predicted image based on a pixel value of a reference image; a prediction error amount deriving unit that derives a prediction error amount that approximates a difference between the image data subjected to the filtering process and the predicted image;

It is configured with! /

[0287] A video decoding device according to the present invention is a video decoding device that decodes encoded data obtained by encoding image data from which frequency components that generate block distortion have been removed.

Variable length decoding means for decoding the encoded data;

Inverse filter processing means for performing inverse filter processing to restore the removed frequency component;

An inverse filter input correcting means for reducing a quantization error occurring in the local decoded image input to the inverse filter processing means;

It is configured with! / [0288] Also, in the video decoding device according to the present invention,

The variable length decoding means decodes a correction method determination result for controlling the correction process of the local decoded image in the inverse filter processing means,

The inverse filter input correction means switches whether to correct the processing target block according to the correction method determination result;

It may be configured as follows.

[0289] Also, in the video decoding device according to the present invention,

The inverse filter input correction means calculates a correction value based on an average pixel value of each of at least one block adjacent to the processing target block, and calculates the correction value and the average pixel value of the processing target block. Switch whether to modify the processing target block according to the comparison result of

It may be configured as follows.

[0290] Also, in the video decoding device according to the present invention,

The inverse filter input correction means calculates a correction value based on an average pixel value of each of at least one block adjacent to the processing target block so that the average pixel value of the processing target block matches the correction value. The processing target block may be corrected.

[0291] The video encoding apparatus according to the present invention divides an image into a plurality of blocks, quantizes the image data of each block, and encodes the quantized representative value obtained by the quantization. Device.

A variable-length encoding unit that encodes information necessary for decoding image data, and a filter processing unit that performs filter processing for removing frequency components that generate block distortion on the image data before quantization. ,

Correction method determining means for determining whether or not to correct the local decoded image in order to reduce the quantization error generated in the local decoded image;

The variable length encoding means encodes a correction method determination result which is a determination result in the correction method determination means;

It may be configured as follows. [0292] Finally, the above-described blocks of the moving image encoding devices 300a and 300b of the present invention and the moving image decoding devices 400a and 400b of the present invention may be configured by hardware logic. It may be realized by software using a CPU as follows.

[0293] That is, the moving picture encoding apparatus and moving picture decoding apparatus include a CPU (central processing unit) that executes instructions of a control program that realizes each function, a ROM (read only memory) that stores the program, and the above It is equipped with RAM (random access memory) for expanding the program, and storage devices (recording media) such as memory for storing the program and various data. An object of the present invention is to compile program codes (execution format program, intermediate code program, source program) of the control program for the video encoding device and the video decoding device, which are software that realizes the above-described functions. A recording medium recorded so as to be readable by a computer is supplied to the moving picture encoding apparatus and the moving picture decoding apparatus, and the computer or CPU or MPU) reads out and executes the program code recorded on the recording medium. This is also achievable.

[0294] Examples of the recording medium include magnetic tapes such as magnetic tapes and cassette tapes, magnetic disks such as floppy (registered trademark) disks / hard disks, and CD-ROM / MO / MD / DVD / CD-R. A disk system including an optical disk, a card system such as an IC card (including a memory card) / optical card, or a semiconductor memory system such as a mask ROM / EPROM / EEPROM / flash ROM can be used.

[0295] Further, the moving image encoding device and the moving image decoding device may be configured to be connectable to a communication network, and the program code may be supplied via the communication network. The communication network is not particularly limited. For example, the Internet, intranet, extranet, LAN, ISDN, VAN, CATV communication network, virtual private network (virtual private network), telephone line network, mobile communication network Satellite communication networks can be used. In addition, the transmission medium constituting the communication network is not particularly limited. For example, in the case of wired communication such as IEEE1394, USB, power line carrier, cable TV line, telephone line, ADSL line, etc., infrared rays such as IrDA remote control, Bluetooth (registered trademark), 802.11 wireless, HDR, mobile phone network, satellite line, terrestrial digital network, etc. can also be used. The present invention is not limited to the above. It can also be realized in the form of a converter data signal embedded in a carrier wave, in which the program code is embodied by electronic transmission.

[0296] The moving picture decoding apparatus according to the present invention divides a quantization target image subjected to filter processing for removing a frequency component that generates block distortion into a plurality of blocks, and the quantization target image is divided into blocks. A video decoding device that inversely quantizes a quantized representative value obtained by quantization, wherein each pixel of a restored quantization target image restored by the above inverse quantization for at least one block A restored quantization target image correcting unit that performs correction by adding an approximate average pixel value that approximates the average pixel value of the quantization target image in the block to which the pixel belongs, and the correction is performed. And an inverse filter processing means for performing an inverse filter process corresponding to the inverse transform of the filter process on the restored quantization target image.

[0297] Also, the moving picture decoding apparatus according to the present invention divides a quantization target image, which has been subjected to filter processing for removing frequency components that generate block distortion, into a plurality of blocks, and quantizes each block for quantization. In order to solve the above-described problem, the image decoding apparatus includes a pixel value of each pixel of the restored quantization target image restored by dequantizing the quantization representative value obtained by the quantization for at least one block. To the restored quantization target image correcting means for performing correction for subtracting the approximate quantization error that approximates the quantization error in the block to which the pixel belongs, and for the restored quantization target image subjected to the above correction, And an inverse filter processing means for performing an inverse filter process corresponding to the inverse transform of the filter process and restoring the removed frequency component.

[0298] Therefore, the block distortion in the decoded moving image can be reduced, and the moving image decoding apparatus does not cause side effects such as blurring of the image, and further generates the removed block distortion. The ability S to achieve a video decoding device that can accurately restore frequency components.

[0299] The specific embodiments or examples made in the detailed description section of the invention have so far clarified the technical contents of the present invention. The present invention should not be construed in a narrow sense, but only by way of example, and can be carried out in various ways within the spirit of the present invention and within the scope of the following claims. . Industrial applicability

The present invention is suitable as a moving image storage device that encodes and stores moving images, a moving image transmission device that encodes and transmits moving images, or a moving image reproducing device that decodes and reproduces moving images. Can be used. Specifically, for example, it can be suitably used for a hard disk recorder or a mobile phone terminal.

Claims

The scope of the claims

[1] A video decoding device that inverse-quantizes the quantized representative value obtained by quantizing the image to be quantized that has been subjected to filter processing to remove frequency components that generate block distortion for each block. There,

For at least one block, an approximate average pixel that approximates the average pixel value of the quantization target image in the block to which the pixel belongs to the pixel value of each pixel of the reconstruction target image restored by the inverse quantization. Reconstructed quantization target image correcting means for performing correction for adding values;

An inverse filter processing means for performing an inverse filter process corresponding to the inverse transform of the filter process on the restored quantization target image subjected to the correction,

A moving picture decoding apparatus characterized by the above.

[2] The restoration quantization target image correction unit further includes correction block designation information acquisition unit that acquires the correction block specification information that specifies the block to be corrected from the moving image encoding device that performs the quantization. Performs the above correction for the block specified by the correction block specification information.

The moving picture decoding apparatus according to claim 1, wherein:

[3] An average pixel value before correction obtained by averaging pixel values of each block of the restored quantization target image, and an average pixel value after correction obtained by subtracting the approximate average pixel value of the block from the average pixel value before correction A comparison estimation means for estimating the magnitude relation of

The restored quantization object image correcting unit performs fiducial correction on the block whose average pixel value after correction is estimated to be smaller than the average pixel value before correction by the comparison and estimation unit! Is,

The moving picture decoding apparatus according to claim 1, wherein:

[4] The moving image according to any one of claims 1 to 3, wherein the approximate average pixel value is derived from the restored quantization target image. Decoding device.

[5] The approximate average pixel value of each block is derived based on the pixel value of the restored quantization target image in the adjacent block for which the approximate average pixel value adjacent to the block has been derived. An approximate average pixel value deriving unit

The approximate average pixel value deriving means includes:

If the correction is not performed for the adjacent block! /, The average pixel value of the restored quantization target image in the adjacent block is set as the approximate average pixel value of the block, and the adjacent block When the above correction is performed, the difference value between the average pixel value of the restored quantization target image in the adjacent block and the approximate average pixel value of the adjacent block is calculated as the approximate average pixel value of the block. To

5. The moving picture decoding apparatus according to claim 4, wherein the moving picture decoding apparatus is provided.

[6] The approximate average pixel value of each block is correlated with the block selected by referring to the quantization parameter that specifies the quantization method, and the approximate average pixel value has been derived. An approximate average pixel value deriving unit for deriving based on the pixel value of the restored quantization target image in the block;

The approximate average pixel value deriving means includes:

If the correction is not performed on the correlation block! /, The average pixel value of the restored quantization target image in the correlation block is set as the approximate average pixel value of the block, and the correlation block When the above correction is performed, the difference value between the average pixel value of the restored quantization target image in the correlation block and the approximate average pixel value of the correlation block is set as the approximate average pixel value of the block.

[7] A video decoding device that inverse-quantizes the quantized representative value obtained by quantizing the image to be quantized that has been subjected to filter processing to remove frequency components that generate block distortion for each block. There,

Correction for subtracting an approximate quantization error that approximates the quantization error in the block to which the pixel belongs from the pixel value of each pixel of the restored quantization image restored by the above inverse quantization for at least one or more blocks And an inverse filter processing means for performing an inverse filter process corresponding to the inverse transform of the filter process on the restored quantization target image subjected to the correction. ,

A moving picture decoding apparatus characterized by the above.

[8] The restored quantization target image correcting unit, further comprising: a corrected block specifying information acquiring unit that acquires the correction block specifying information that specifies the block to be corrected from the moving image encoding device that performs the quantization. Performs the above correction for the block specified by the correction block specification information.

8. The moving picture decoding apparatus according to claim 7, wherein

[9] The approximate quantization error in each block is

A difference value between a weighted average value obtained by weighted average of the average pixel values of the restored quantization target image in each block adjacent to the block and an average pixel value of the restored quantization target image in the block;

9. The moving picture decoding apparatus according to claim 7, wherein the moving picture decoding apparatus is characterized in that:

[10] The filter inverse processing means includes:

An average pixel value calculation process for calculating an average pixel value in each block of the processing target image to be subjected to the inverse filtering process;

A prediction process for predicting a pixel value of each pixel of the processing target image based on the average pixel value calculated in the average pixel value calculation process;

For each pixel of the processing target image, the difference between the predicted pixel value of the pixel predicted in the prediction process and the average pixel value of the block to which the pixel belongs calculated in the average pixel value calculation process Is an additional component to be added to the processing target image, and the addition processing is performed to add the additional component to the pixel value of the pixel of the processing target image.

The moving picture decoding apparatus according to any one of claims 1 to 9, wherein the moving picture decoding apparatus is characterized in that:

[11] A moving picture encoding apparatus that divides a quantization target image into a plurality of blocks and quantizes each block,

Filter processing means for performing filtering processing on the quantization target image to remove frequency components that generate block distortion;

For at least one block, the quantization target image in the block to which the pixel belongs is calculated from the pixel value of each pixel of the quantization target image subjected to the filtering process. A quantization target image correcting unit that performs correction to subtract an approximate average pixel value that approximates an average pixel value of an image, and

A moving picture coding apparatus characterized by the above.

[12] For each block of the quantization target image, an average pixel value before correction obtained by averaging pixel values of the quantization target image in the block, and the approximate average pixel of the block from the average pixel value before correction Average pixel value calculating means for calculating a corrected average pixel value obtained by subtracting the value;

Correction evaluation means for comparing the size of the average pixel value before correction and the average pixel value after correction for each block of the quantization target image,

The quantization target image correcting unit applies the correction to the block in which the average pixel value after correction is determined to be smaller than the average pixel value before correction by the correction evaluation unit! Is,

12. The moving picture coding apparatus according to claim 11, wherein the moving picture coding apparatus is provided.

[13] Correction block designation information for designating a block that has been corrected by the quantization target image correction means for a video decoding device that inversely quantizes the quantized representative value obtained by the quantization 13. The moving picture coding apparatus according to claim 11, further comprising correction block designation information providing means for providing

[14] A moving picture coding apparatus that divides a coding target picture subjected to filter processing for removing frequency components that generate block distortion by a filter processing means into a plurality of blocks, and codes each block.

A local decoding means for decoding the encoded data obtained by the above encoding and obtaining a locally decoded image;

For each block, any one of a weighted average value obtained by weighted average of the average pixel values of the local decoded image in each block adjacent to the block and an average pixel value of the local decoded image in the block Correction method determining means for determining whether a value is close to an average pixel value of the encoding target image in the block;

A block in which the weighted average value is determined to be closer to the average pixel value of the encoding target image with respect to the video decoding device that decodes the encoded data obtained by the encoding. Correction block instruction information providing means for providing correction block instruction information for instructing as a block to be corrected,

A moving picture coding apparatus characterized by the above.

[15] The filtering means includes:

An average pixel value calculation process for calculating an average pixel value in each block of the processing target image to be filtered;

For each pixel of the processing target image, the difference between the predicted pixel value of the pixel predicted in the prediction process and the average pixel value of the block to which the pixel belongs calculated in the average pixel value calculation process Is a removal component to be removed from the processing target image, and a subtraction process for subtracting the removal component from the pixel value of the pixel of the processing target image, and

15. The moving picture encoding apparatus according to claim 1, wherein any one of claims 11 to 14 is used.

[16] A moving picture reproduction apparatus comprising the moving picture decoding apparatus according to any one of claims 1 to 10, wherein the moving picture decoded by the moving picture decoding apparatus is A video playback device for playback.

[17] A moving picture recording apparatus comprising the moving picture encoding apparatus according to any one of claims 11 to 15, encoded by the moving picture encoding apparatus A moving image recording device that records moving images.