JPH06334983A - Encoding method, encoding device, decoding method, and decoding device for picture signal - Google Patents

Abstract

PURPOSE:To quantize or inversely quantize picture data with the quantization characteristic of proper precision in a wide range without increasing the scale of a constitution circuit by performing quantization based on the nonlinear quantization characteristic expressed with the multiplication value between the value of a power of two and its coefficient. CONSTITUTION:A quantization characteristic QUANT is obtained in accordance with a specific formula where (k) is first quantization information indicating the exponent of the power of two and i/2+j is second quantization information indicating the coefficient by which the value of the power of two should be multiplied. (j) and (k) are positive integers, and (i) is 0 or one. When quantization information consists of 5 bits and bits are denoted as Q1 to Q5 in order from the most significant bit, (k), (j), and (i) are expressed by (Q1Q2), (Q3Q4), and (Q5) respectively. In the quantization characteristic reproducing circuit shown in the figure, the value of quantization information (Q3Q4Q5) is inputted to a shifter 110, and the value of quantization information (Q1Q2) is used to shift it, and the 7-bit quantization characteristic QUANT is finally reproduced from the output obtained by conversion of quantization information (Q1Q2) in a table part 111.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention uses a method of quantizing and dequantizing image data using a storage system recording medium such as an optical disk or a magnetic tape, and a storage system recording medium such as an optical disk or a magnetic tape. Information recording device and information reproducing device,
Further, the present invention relates to an information transmitting apparatus and an information receiving apparatus suitable for, for example, a so-called video conference system, video telephone system, and broadcasting equipment.

[0002]

2. Description of the Related Art Recently, in a so-called signal transmission system for transmitting video signals and audio signals to a remote place, such as a video conference system and a video telephone system,
In order to use the transmission path efficiently, it has been attempted to improve the transmission efficiency of information by encoding video signals and audio signals.

In particular, since moving image data has an extremely large amount of information, when this information is recorded for a long time, the video signal is highly efficient coded and recorded, and the recorded signal is efficiently read. A means for good decoding is indispensable, and in order to meet such a demand, a high-efficiency coding method utilizing the correlation of video signals has been proposed. One of the high-efficiency coding methods is MPEG (Moving Picture Expert).
s Group) method.

This MPEG system first reduces the redundancy in the time axis direction by taking the difference between the image frames of the video signal by utilizing the inter-frame correlation, and then by using the line correlation, the discrete cosine is used. The video signal is efficiently coded by reducing the redundancy in the spatial axis direction by using processing such as conversion (DCT).

When interframe correlation is used, for example, FIG.
As shown in (A) of FIG. ₁, T₂, T₃To
And the frame images PC1, PC2, and PC3 are
When it occurs, the image of frame image PC1 and PC2
The difference between the image signals is calculated and the image is displayed as shown in FIG.
Generate the PC 12 and also the frame image of FIG.
By calculating the difference between the image signals of PC2 and PC3,
The image PC 23 of (B) is generated. Usually adjacent in time
Since the frame image to be changed does not change so much,
The difference signal when calculating the difference between two frame images is small
It will be a small value.

That is, the image PC1 shown in FIG. 9B.
2, the frame images PC1 and P of FIG.
As a difference between the image signals of C2, the image PC1 of FIG.
2, the difference signal of the shaded portion in the figure is obtained, and in the image PC23 shown in FIG. 9B, the difference between the image signals of the frame images PC2 and PC3 of FIG. The difference signal of the shaded portion in the image PC23 of FIG. 9B is obtained. Therefore, if this difference signal is encoded, the code amount can be compressed.

However, since the original image cannot be restored by transmitting only the difference signal, the image of each frame is converted into an I picture (Intra-coded picture: intra-coded or intra-coded image). P picture (Pre
dictive-coded picture: Bi-directionally predictive-coded picture:
The image signal is compressed and coded as one of the pictures (bidirectional predictive coded image).

That is, for example, (A) and (B) of FIG.
As shown in, the image signals of 17 frames from frame F1 to frame F17 are group of pictures,
One unit of processing. Then, the leading frame F1
The image signal of is encoded as an I picture, the second frame F2 is processed as a B picture, and the third frame F3 is processed as a P picture. Less than,
The fourth and subsequent frames F4 to F17 are B
Alternately processed as picture or P picture.

As the image signal of the I picture, the image signal for one frame is transmitted as it is. On the other hand, as the image signal of the P picture, basically, as shown in FIG.
As shown in (A) of 0, I which precedes it by I
The difference from the picture signal of the picture or P picture is encoded and transmitted. Further, as the image signal of the B picture, basically, as shown in FIG. 10B, the difference from the average value of the image signals of both the temporally preceding frame and the temporally following frame is obtained. , The difference is encoded and transmitted.

FIGS. 11A and 11B show the principle of the method of encoding a moving image signal in this way. Note that FIG. 11A schematically shows frame data of a moving image signal, and FIG. 11B schematically shows transmitted frame data. As shown in FIG. 11, since the first frame F1 is processed as an I picture, that is, a non-interpolation frame, it is directly transmitted to the transmission path as the transmission data F1X (transmission non-interpolation frame data) (intra-picture coding). . On the other hand, since the second frame F2 is processed as a B picture, that is, an interpolation frame,
The frame F1 preceding in time and the frame F3 following in time (non-interpolation frame of interframe coding)
Is calculated and the difference is transmitted as transmission data (transmission interpolation frame data) F2X.

However, there are four types of processing as the B picture, which will be described in more detail. In the first process, the data of the original frame F2 is converted into an arrow SP with a broken line in the figure.
As shown by 1, the data is transmitted as it is as the transmission data F2X (intra coding), and the processing is the same as in the case of the I picture. The second process is to calculate a difference from the frame F3 that is temporally following and transmit the difference as indicated by a dashed arrow SP2 in the figure (backward predictive coding). The third process is the dashed arrow S in the figure.
As indicated by P3, the difference from the frame F1 preceding in time is transmitted (forward predictive coding). Further, in the fourth processing, as shown by a broken line arrow SP4 in the figure, a difference between the temporally preceding frame F1 and the average value of the following frame F3 is generated, and this difference is transmitted data F
It is transmitted as 2X (bidirectional predictive coding).

Of these four methods, the method that minimizes the amount of transmitted data is adopted.

When transmitting the difference data, the motion vector x1 between the image of the frame for which the difference is to be calculated (predicted image) (the motion between the frames F1 and F2 in the case of forward predictive coding). Vector) or motion vector x2 (frames F3 and F in the case of backward prediction coding)
2), or motion vectors x1 and x
Both of the two (for bidirectional prediction) are transmitted with the difference data.

A frame F3 (non-interpolation frame for inter-frame coding) of a P picture has a temporally preceding frame F1 as a prediction image and a difference signal (shown by a broken line arrow SP3) from the frame F1. Motion vector x3
Is calculated and transmitted as transmission data F3X (forward predictive coding). Alternatively, the original frame F3
Data is transmitted as it is as transmission data F3X (indicated by a dashed arrow SP1) (intra-encoding). This P
Which method is used to transmit a picture is the same as in the case of a B picture, and the one with less transmission data is selected.

The B-picture frame F4 and the P-picture frame F5 are processed in the same manner as described above, and the transmission data F4X, F5X, the motion vectors x4, x5, x6.
Etc. are obtained.

FIG. 12 is a diagram showing another example of a method for intra-frame / inter-frame coding of an image sequence. In FIG. 12, the cycle of 15 frames is one unit of encoding.

Here, frame 2 is an I-picture which is intra-frame coded, and frames 5, 8, 11, 14 are
Is a P picture that is predicted only from the forward direction and is inter-frame coded, and frames 0, 1, 3, 4, 6,
7, 9, 10, 12, and 13 are B pictures that are predicted from both the forward and backward directions and are inter-frame coded.

FIG. 13 shows an input order, an encoding order, a decoding order, and an output (display) order in this intra-frame / inter-frame coding.

FIG. 14 shows an example of the configuration of an apparatus for encoding and transmitting a moving image signal based on the above-mentioned principle, and for decoding this. The encoding device 1 is configured to encode the input video signal, transmit it to the recording medium 3 as a transmission path, and record it. Then, the decoding device 2
The signal recorded on the recording medium 3 is reproduced, and this is decoded and output.

First, in the encoding device 1, the video signal VD input via the input terminal 10 is processed by the preprocessing circuit 11.
To the luminance and chrominance signals (in this example,
Color difference signals) are separated and A / D converted by A / D (analog / digital) converters 12 and 13, respectively. A /
The video signal converted into a digital signal by A / D conversion by the D converters 12 and 13 is sent to and stored in the frame memory 14. In the frame memory 14, the luminance signal is stored in the luminance signal frame memory 15, and the color difference signal is stored in the color difference signal frame memory 16, respectively.

Next, the format conversion circuit 17 converts the frame format signal stored in the frame memory 14 into a block format signal. That is,
As shown in (A) of FIG. 15, the video signal stored in the frame memory 14 is frame format data in which V dots of H dots per line are collected. The format conversion circuit 17 divides this 1-frame signal into N slices in units of 16 lines. Then, each slice is divided into M macroblocks, as shown in FIG. Each macroblock has a size of 16 × 1 as shown in FIG.
It is composed of a luminance signal corresponding to 6 pixels (dots), and this luminance signal further includes blocks Y [1] to Y in units of 8 × 8 dots, as shown in FIG.
It is classified into [4]. The 16 × 16 dot luminance signal corresponds to the 8 × 8 dot Cb signal and the 8 × 8 dot Cr signal.

At this time, the arrangement of the moving picture signals in each slice shown in FIG. 15A is such that the moving picture signals are continuous in macroblock units shown in FIG. 15C. Within a macro block, moving image signals are made continuous in units of minute blocks in the order of raster scanning.

The data converted into the block format as described above is supplied from the format conversion circuit 17 to the encoder 18 and encoded (encoded). Details of the encoder 18 will be described later with reference to FIG.

The signal encoded by the encoder 18 is output to the transmission path as a bit stream and recorded on the recording medium 3, for example. The data reproduced from the recording medium 3 is supplied to the decoder 31 of the decoding device 2 and decoded (decoded). Details of the decoder 31 will be described later with reference to FIG.

The data decoded by the decoder 31 is input to the format conversion circuit 32 and converted from the block format to the frame format. The luminance signal of this frame format is stored in the luminance signal frame memory 34 of the frame memory 33.
Is sent to and stored in the color difference signal frame memory 35. The luminance signal and the color difference signal read from the luminance signal frame memory 34 and the color difference signal frame memory 35 are respectively D / A converted by the D / A converters 36 and 37, and further supplied to the post-processing circuit 38. The post-processing circuit 38 combines them. The output video signal is output from the output terminal 30 to a display (not shown), such as a CRT, for display.

Next, the structure of the encoder 18 will be described with reference to FIG.

First, the image data to be encoded, which is supplied through the input terminal 49, is input to the motion vector detection circuit 50 in units of the macro blocks. The motion vector detection circuit 50 converts the image data of each frame into an I-picture, according to a preset predetermined sequence.
It is processed as a P picture or a B picture. here,
The image of each frame that is sequentially input is
Which of P and B pictures is to be processed is predetermined (for example, as shown in FIG. 10, the group of pictures composed of frames F1 to F17 is I, B, P, B, P, ... B, P).

The image data of the frame (for example, the frame F1) processed as the I picture is transferred from the motion vector detection circuit 50 to the front original image portion 51a of the frame memory 51 and stored therein, and is processed as the B picture. Image data of a frame (for example, frame F2) to be stored in the original image portion (reference original image portion) 51b is stored, and image data of a frame (for example, frame F3) processed as a P picture is rearward original image portion. It is transferred to 51c and stored.

At the next timing, B
When an image of a frame to be processed as a picture (for example, the frame F4) or a P picture (for example, frame F5) is input, the first P picture (frame F3) stored in the backward original image portion 51c until then is input. The image data is transferred to the front original image portion 51a, and the image data of the next B picture (frame F4) is transferred to the original image portion 51b.
Is stored (overwritten) in the next P picture (frame F
The image data of 5) is stored (overwritten) in the rear original image portion 51c. Such an operation is sequentially repeated.

The signal of each picture stored in the frame memory 51 is read out and sent to the prediction mode switching circuit 52. In this prediction mode switching circuit 52, frame prediction mode processing or field prediction mode processing is performed. Furthermore, under the control of the prediction determination circuit 54, the calculation unit 53 performs calculation of intra-picture prediction, forward prediction, backward prediction, or bidirectional prediction. Which of these processes is to be performed is determined in accordance with the prediction error signal (the difference between the reference image to be processed and the predicted image for this). For this reason,
The motion vector detection circuit 50 generates the sum of absolute values (may be the sum of squares) of the prediction error signal used for this determination.

Here, the frame prediction mode and field prediction mode in the prediction mode switching circuit 52 will be described.

When the frame prediction mode is set in the prediction mode switching circuit 52, the prediction mode switching circuit 52 supplies the four luminance blocks Y [1] to Y [4] supplied from the motion vector detection circuit 50. ],
It is output as it is to the arithmetic unit 53 in the subsequent stage. That is, in this case, as shown in FIG. 17A, the data of the odd field lines and the data of the even field lines are mixed in each luminance block.
The solid line in each macroblock in FIG. 17 indicates the data of the line of the odd field (the line of the first field), and the broken line indicates the data of the line of the even field (the line of the second field). Symbols b and b indicate units of motion compensation. In the frame prediction mode, prediction is performed in units of four luminance blocks (macro blocks), and one motion vector is associated with each of the four luminance blocks.

On the other hand, the prediction mode switching circuit 5
When the field prediction mode is set in 2, the signal input from the motion vector detection circuit 50 in the configuration shown in FIG. 17A is converted into, for example, four signals as shown in FIG. 17B. Luminance block Y of luminance blocks
[1] and Y [2] are composed of only the data of the lines in the odd field, and the other two luminance blocks Y [3] and Y [3]
[4] is composed only of the data of the lines of the even fields and is output to the arithmetic unit 53. In this case,
One motion vector corresponds to two luminance blocks Y [1] and Y [2], and the other two luminance blocks Y
Another one motion vector is associated with [3] and Y [4].

Describing according to the configuration of FIG. 16, the motion vector detection circuit 50 calculates the sum of absolute values of prediction errors in the frame prediction mode and the sum of absolute values of prediction errors in the field prediction mode into the prediction mode switching circuit 52. Output to. The prediction mode switching circuit 52 compares the absolute value sums of the prediction errors in the frame prediction mode and the field prediction mode, performs the above-described processing corresponding to the prediction mode with the smaller value, and obtains the result. The data is output to the calculation unit 53.

However, such processing is actually performed by the motion vector detection circuit 50. That is, the motion vector detection circuit 50 outputs a signal having a configuration corresponding to the determined mode to the prediction mode switching circuit 52, and the prediction mode switching circuit 52 outputs the signal as it is to the arithmetic unit 53 in the subsequent stage. .

In the case of the frame prediction mode, the color difference signal, as shown in FIG. 17 (A), is in the state where the data of the odd field lines and the data of the even field lines are mixed, and the arithmetic unit 53 is used. Is supplied to. Further, in the case of the field prediction mode, as shown in FIG. 17B, the upper half (4 lines) of each color difference block Cb [5], Cr [6] is set to the luminance blocks Y [1], Y [. 2] and the lower half (4 lines) of each color difference block Cb [5], Cr [6] is an even field corresponding to a luminance block Y [3], Y [4]. Color difference signal.

Further, the motion vector detection circuit 50 uses the prediction determination circuit 54 in the intra-picture prediction,
A sum of absolute values of prediction errors for determining whether to perform forward prediction, backward prediction, or bidirectional prediction is generated.

That is, the sum ΣAij of the signals Aij of the macroblocks of the reference image is used as the sum of the absolute values of the prediction errors of the intra-picture prediction.
The absolute value | ΣAij | of the macroblock and the sum Σ | Aij | of the absolute values | Aij | of the signal Aij of the macroblock are calculated. Also, as the sum of absolute values of prediction errors in forward prediction, the sum Σ of absolute values | Aij-Bij | of the difference (Aij-Bij) between the signal Aij of the macroblock of the reference image and the signal Bij of the macroblock of the predicted image. | Aij-Bij | is calculated. Further, the sum of absolute values of the prediction errors of the backward prediction and the bidirectional prediction is also obtained in the same manner as in the case of forward prediction (the predicted image is changed to a predicted image different from that in forward prediction).

The sum of these absolute values is supplied to the prediction judgment circuit 54. The prediction determination circuit 54 selects the smallest sum of absolute values of prediction errors of forward prediction, backward prediction, and bidirectional prediction as the sum of absolute values of prediction errors of inter prediction. Further, the absolute value sum of the prediction error of the inter prediction is compared with the absolute value sum of the prediction error of the intra-picture prediction, and the smaller one is selected, and the mode corresponding to the selected absolute value sum is selected as the prediction mode. To do. That is, if the sum of absolute values of prediction errors in intra-picture prediction is smaller, the intra-picture prediction mode is set. If the sum of the absolute values of the prediction errors of the inter prediction is smaller, the mode having the smallest sum of the corresponding absolute values is set among the forward prediction, the backward prediction, and the bidirectional prediction mode.

In this way, the motion vector detection circuit 50
Is a configuration in which the signal of the macroblock of the reference image corresponds to the frame prediction mode or field prediction mode selected by the prediction mode switching circuit 52 as shown in FIG. A motion vector between the prediction image corresponding to the prediction mode selected by the prediction determination circuit 54 among the four prediction modes and the reference image is detected, and the variable length coding circuit 58 described later is provided. It is output to the motion compensation circuit 64. As described above, the motion vector that minimizes the sum of the absolute values of the corresponding prediction errors is selected.

When the motion vector detection circuit 50 reads the image data of the I picture from the front original image portion 51a, the prediction determination circuit 54 sets the intra-frame (image) prediction mode (mode without motion compensation) as the prediction mode. After setting, the switch of the calculation unit 53 is switched to the contact a side. As a result, the I-picture image data is input to the DCT mode switching circuit 55.

The DCT mode switching circuit 55, as shown in FIG. 18A or 18B, is a state in which the data of four luminance blocks are mixed with the lines of the odd field and the lines of the even field ( Frame DCT mode), or a state in which odd field lines and even field lines are separated (field DCT mode)
And output to the DCT circuit 56.
That is, the DCT mode switching circuit 55 mixes odd field data and even field data in a DC
The coding efficiency in the case where the T processing is performed is compared with the coding efficiency in the case where the data in the odd field and the data in the even field are separated and subjected to the DCT processing, and a mode having a good coding efficiency is selected.

For example, as shown in (A) of FIG. 18, the input signal has a structure in which lines of odd fields and lines of even fields are mixed, and signals of lines of odd fields vertically adjacent to each other are even. The difference from the signal on the field line is calculated, and the sum (or sum of squares) of the absolute values is calculated. In addition, the input signal is shown in FIG.
As shown in (B), the odd field line and the even field line are separated, and the difference between the signals of the vertically adjacent odd field lines and the signal between the even field lines is calculated. Each is calculated and the sum (or sum of squares) of each absolute value is obtained. further,
The two obtained (absolute value sum) are compared, and the DCT mode corresponding to a smaller value is set. That is, the DCT mode switching circuit 55 sets the frame DCT mode if the former is smaller, and sets the field DCT mode if the latter is smaller. And the selected DCT
The data having the configuration corresponding to the mode is output to the DCT circuit 56, and the DCT flag indicating the selected DCT mode is output to the variable length coding circuit 58 and the motion compensation circuit 64.

The prediction mode in the prediction mode switching circuit 52 (see FIG. 17) and the DCT mode switching circuit 5
As is clear from a comparison with the DCT mode in FIG. 5 (see FIG. 18), the data structure in each mode is substantially the same for the luminance block.

When the frame prediction mode (a mode in which odd lines and even lines are mixed) is selected in the prediction mode switching circuit 52, the frame DCT mode (in which odd lines and even lines are mixed) is also selected in the DCT mode switching circuit 55. If the field prediction mode (mode in which data of odd field and data of even field are separated) is selected in the prediction mode switching circuit 52,
Field DC in the DCT mode switching circuit 55
There is a high possibility that the T mode (mode in which the data in the odd field and the data in the even field are separated) is selected.

However, this is not always the case, and the prediction mode switching circuit 52 determines the prediction mode so that the sum of the absolute values of the prediction errors becomes smaller, and the DCT mode switching circuit 55 determines the sign. The DCT mode is determined so that the conversion efficiency is good.

The I-picture image data output from the DCT mode switching circuit 55 is input to the DCT circuit 56, subjected to DCT (discrete cosine transform) processing, and converted into DCT coefficients. The DCT coefficient is input to the quantization circuit 57, quantized in a quantization step corresponding to the data storage amount (buffer storage amount) of the transmission buffer 59 in the subsequent stage, and then input to the variable length coding circuit 58.

The variable length coding circuit 58 is a quantization circuit 57.
The image data (in this case, I picture data) supplied from the quantization circuit 57 corresponding to the supplied quantization step (scale) is, for example, Huffman (Huffm).
an) Converted into a variable length code such as code, and transmitted to the transmission buffer 59
Output to. Further, the variable length coding circuit 58 is set with a quantization step (scale) from the quantization circuit 57 and a prediction mode from the prediction determination circuit 54 (either intra-picture prediction, forward prediction, backward prediction, or bidirectional prediction). The motion vector from the motion vector detection circuit 50, the prediction flag from the prediction mode switching circuit 52 (flag indicating whether the frame prediction mode or the field prediction mode has been set), and the DCT mode switching circuit. A DCT flag (flag indicating whether the frame DCT mode or the field DCT mode is set) is input from 55, and these are also variable length coded.

The transmission buffer 59 temporarily stores the input data and stores the data corresponding to the storage amount in the quantizing circuit 57.
Output to. When the remaining amount of data in the transmission buffer 59 increases to the allowable upper limit value, the transmission buffer 59 reduces the data amount of the quantized data by increasing the quantization step of the quantization circuit 57 by the quantization control signal. On the contrary, when the remaining amount of data in the transmission buffer 59 decreases to the allowable lower limit value, the transmission buffer 59 reduces the quantization step of the quantization circuit 57 by the quantization control signal, and thus the quantized data is reduced. Increase the amount of data. In this way, overflow or underflow of the transmission buffer 59 is prevented. Then, the data accumulated in the transmission buffer 59 is read out at a predetermined timing, is output to the transmission line via the output terminal 69, and is recorded on the recording medium 3, for example.

On the other hand, the I picture data output from the quantizing circuit 57 is input to the inverse quantizing circuit 60 and inversely quantized corresponding to the quantizing step supplied from the quantizing circuit 57. The output of the inverse quantization circuit 60 is IDCT (inverse D
(CT) circuit 61 and is subjected to inverse DCT processing.

Here, the prediction flag from the prediction mode switching circuit 52 and the DCT flag from the DCT mode switching circuit 55 are input to the conversion circuit 66. Further, the prediction flag from the prediction mode switching circuit 52 is input to the conversion circuit 65. The data subjected to the inverse DCT processing by the IDCT circuit 61 is converted into a conversion circuit 66 and an arithmetic unit 62.
And the conversion circuit 65, the data is matched, and then the forward prediction image portion 63 of the frame memory 63 is obtained.
is supplied to and stored in a.

By the way, the motion vector detection circuit 50
When the image data of each frame that is sequentially input is processed as a picture of, for example, I, B, P, B, P, B ..., The image data of the first input frame is an I picture. , The image data of the next input frame is processed as a P picture before the image of the next input frame is processed as a B picture. That is, since the B picture is accompanied by backward prediction, it cannot be decoded unless the P picture as the backward predicted image is prepared in advance.

Therefore, the motion vector detection circuit 50
After the picture processing, the processing of the picture data of the P picture stored in the backward original image section 51c is started. Then, as in the case described above, the sum of absolute values of the inter-frame difference (prediction error) in macroblock units is supplied from the motion vector detection circuit 50 to the prediction mode switching circuit 52 and the prediction determination circuit 54. Prediction mode switching circuit 5
2 and the prediction determination circuit 54 sets the frame / field prediction mode or the prediction mode of intra-picture prediction, forward prediction, backward prediction, or bidirectional prediction in accordance with the sum of absolute values of prediction errors of macroblocks of the P picture. Set.

When the intra-picture prediction mode is set,
The switch in the calculation unit 53 is switched to the contact a side as described above. Therefore, the image data of the P picture is I
Similar to the picture image data, it is transmitted to the transmission line via the DCT mode switching circuit 55, the DCT circuit 56, the quantization circuit 57, the variable length coding circuit 58, and the transmission buffer 59. In addition, the image data is the inverse quantization circuit 60,
The backward predicted image section 6 of the frame memory 63 via the IDCT circuit 61, the conversion circuit 66, the computing unit 62, and the conversion circuit 65.
3b and stored.

On the other hand, in the forward prediction mode, the switch in the arithmetic unit 53 is switched to the contact b side, and
The image data (in this case, an image of an I picture) stored in the forward prediction image portion 63a of the frame memory 63 is read out, and the motion compensation circuit 64 corresponds to the motion vector output by the motion vector detection circuit 50. Motion compensated. That is, the motion compensation circuit 64 includes the prediction determination circuit 54.
When the forward prediction mode setting is instructed, the read address of the forward prediction image unit 63a is shifted from the position corresponding to the position of the macro block currently output by the motion vector detection circuit 50 by the amount corresponding to the motion vector. Data is read out to generate predicted image data.

The predicted image data output from the motion compensation circuit 64 is supplied to the calculator 53a. Calculator 53a
Subtracts the predicted image data corresponding to the macroblock supplied from the motion compensation circuit 64 from the macroblock data of the reference image supplied from the prediction mode switching circuit 52, and outputs the difference (prediction error). To do. This difference data is the DCT mode switching circuit 55, DC
T circuit 56, quantization circuit 57, variable length coding circuit 58,
It is transmitted from the output terminal 69 to the transmission line via the transmission buffer 59. In addition, this difference data is stored in the inverse quantization circuit 60.
And is locally decoded by the IDCT circuit 61 and input to the computing unit 62 via the conversion circuit 66.

The conversion circuit 66 is supplied with the prediction flag from the prediction mode switching circuit 52 and the DCT flag from the DCT mode switching circuit 55, whereby the output from the IDCT circuit 61 is matched. .

Further, the arithmetic unit 62 includes an arithmetic unit 53a.
The same data as the predicted image data supplied to the computer is supplied. The calculator 62 adds the predicted image data output by the motion compensation circuit 64 to the difference data output by the IDCT circuit 61. As a result, the image data of the original (decoded) P picture is obtained. The image data of the P picture is supplied to the backward predicted image portion 63b of the frame memory 63 via the conversion circuit 65 and stored therein.

In this way, after the I picture data and the P picture data are stored in the forward prediction image portion 63a and the backward prediction image portion 63b, respectively, the motion vector detection circuit 50 then processes the B picture. To execute. The prediction mode switching circuit 52 sets the frame mode or the field mode in accordance with the magnitude of the sum of absolute values of the inter-frame differences in macroblock units, and the prediction determination circuit 54 sets the prediction mode to the intra-picture prediction mode. It is set to either the forward prediction mode, the backward prediction mode, or the bidirectional prediction mode.

As described above, in the intra-frame prediction mode or the forward prediction mode, the switch in the arithmetic unit 53 is switched to the contact a or the contact b. At this time, the same processing as in the P picture is performed and the data is transmitted. On the other hand, when the backward prediction mode or the bidirectional prediction mode is set, the switch in the calculation unit 53 is switched to the contact c or the contact d.

In the backward prediction mode in which the switch in the arithmetic unit 53 is switched to the contact c side, the image data (in this case, the image of the P picture) stored in the backward predicted image unit 63b is read out and moves. The compensation circuit 64 compensates for the motion vector output from the motion vector detection circuit 50. That is, the motion compensation circuit 64
When the prediction determination circuit 54 instructs to set the backward prediction mode, the read address of the backward predicted image portion 63b is set to the motion vector from the position corresponding to the position of the macro block currently output by the motion vector detection circuit 50. The data is read out by shifting by the amount corresponding to, and predicted image data is generated.

The predicted image data output from the motion compensation circuit 64 is supplied to the calculator 53b. Calculator 53b
Subtracts the predicted image data supplied from the motion compensation circuit 64 from the data of the macroblock of the reference image supplied from the prediction mode switching circuit 52, and outputs the difference. This difference data is stored in the DCT mode switching circuit 5
5, the DCT circuit 56, the quantization circuit 57, the variable length coding circuit 58, and the transmission buffer 59, and the data is transmitted from the output terminal 69 to the transmission line.

In the bidirectional prediction mode in which the switch in the calculation unit 53 is switched to the contact point d side, the image (in this case, I picture image) data stored in the forward prediction image unit 63a and the backward prediction image unit are stored. The image data (image of P picture in this case) stored in 63b is read out, and the motion compensation circuit 64 performs motion compensation corresponding to the motion vector output from the motion vector detection circuit 50. That is, the motion compensation circuit 64 includes the prediction determination circuit 5
4 is instructed to set the bidirectional prediction mode, the read addresses of the forward prediction image portion 63a and the backward prediction image portion 63b are moved from the position corresponding to the position of the macro block currently output by the motion vector detection circuit 50. The data is read out by shifting by the amount corresponding to the vector (the motion vector in this case is two for the forward prediction image and the backward prediction image), and predicted image data is generated.

The predicted image data output from the motion compensation circuit 64 is supplied to the calculator 53c. Calculator 53c
Subtracts the average value of the predicted image data supplied from the motion compensation circuit 64 from the macroblock data of the reference image supplied from the motion vector detection circuit 50, and outputs the difference. This difference data is output through a DCT mode switching circuit 55, a DCT circuit 56, a quantization circuit 57, a variable length coding circuit 58, and a transmission buffer 59 to an output terminal 69.
Is transmitted to the transmission line.

The B picture image is not stored in the frame memory 63 because it is not used as a predicted image of another image.

In the frame memory 63, the forward predicted image portion 63a and the backward predicted image portion 63b are bank-switched as necessary, and are stored in one or the other with respect to a predetermined reference image. It is possible to switch and output an image as a forward prediction image or a backward prediction image.

In the above description, the description has been centered on the luminance block, but the same applies to the color difference block in FIG.
7 and the macroblocks shown in FIG. 18 are processed as a unit and transmitted. The motion vector used for processing the color difference block is obtained by halving the motion vector of the corresponding luminance block in each of the vertical direction and the horizontal direction.

Next, FIG. 19 is a block diagram showing an example of the configuration of the decoder 31 shown in FIG. The coded image data transmitted through the transmission path (recording medium 3) is received by a receiving circuit (not shown) or reproduced by a reproducing device, and then is received by the receiving buffer 81 via the input terminal 80.
Is temporarily stored in. After that, the temporarily stored image data is supplied to the variable length decoding circuit 82 of the decoding circuit 90.
The variable length decoding circuit 82 performs variable length decoding on the data supplied from the reception buffer 81, and calculates the motion vector, the prediction mode, the prediction flag and the DCT flag by the motion compensation circuit 87.
In addition, the quantization step is output to the inverse quantization circuit 83, and the decoded image data is output to the inverse quantization circuit 83.

The inverse quantization circuit 83 is a variable length decoding circuit 8
The image data supplied from No. 2 is inversely quantized in accordance with the supplied quantization step and output to the IDCT circuit 84. Data output from the inverse quantization circuit 83 (DCT
The coefficient) is subjected to inverse DCT processing by the IDCT circuit 84 and supplied to the calculator 85 via the conversion circuit 88.

The conversion circuit 88 is supplied with the prediction flag and the DCT flag, and the conversion circuit 88 takes the consistency of the image data supplied from the IDCT circuit 84 based on these flags.

When the image data supplied to the calculator 85 is I-picture data, the data is calculated by the calculator 8
5 is supplied to the forward predictive image section 86a in the frame memory 86 via the conversion circuit 89 in order to generate predictive image data of the image data (P or B picture data) output from the computer 5 and input to the calculator 85 later. Are stored. In addition, this data is output to the output terminal 91 via the conversion circuit 89.
It is output to the format conversion circuit 32 of FIG.

The image data supplied to the arithmetic unit 85 has the image data one frame before as the predicted image data P.
When the data is picture data and is data in the forward prediction mode, the forward prediction image section 8 of the frame memory 86 is used.
Image data (I
(Picture data) is read out, and the motion compensation circuit 87 performs motion compensation corresponding to the motion vector output from the variable length decoding circuit 82. Then, in the calculator 85, the image data (difference data) supplied from the IDCT circuit 84 via the conversion circuit 88 is added and output. This added data, that is, the decoded P picture data, is passed through the conversion circuit 89 to generate predictive image data of the image data (B picture or P picture data) input to the calculator 85 later. And is supplied to and stored in the backward predicted image portion 86b in the frame memory 86.

Even in the case of P-picture data, the data in the intra-picture prediction mode is not processed by the arithmetic unit 85 as in the I-picture data, and the conversion circuit 89 is used as it is.
Is stored in the backward predicted image portion 86b via the. Since this P picture is an image to be displayed next to the next B picture, the format conversion circuit 3 is still at this point.
2 is not output (as described above, the P picture input after the B picture is processed and transmitted before the B picture).

When the image data supplied from the IDCT circuit 84 is B-picture data, the forward prediction image portion in the frame memory 86 corresponds to the prediction mode supplied from the variable length decoding circuit 82. The image data of the I picture (in the forward prediction mode) stored in 86a, the image data of the P picture in the backward prediction image section 86b (in the backward prediction mode), or both of them (in both directions). (In the case of the prediction mode) is read, and in the motion compensation circuit 87, the variable length decoding circuit 82
The motion compensation corresponding to the output motion vector is performed, and the predicted image is generated. However, when motion compensation is not required (in the case of the intra-picture prediction mode), the predicted picture is not generated.

The data thus motion-compensated by the motion compensation circuit 87 is added to the output from the conversion circuit 88 in the calculator 85. This addition output is output from the output terminal 91 to the format conversion circuit 32 of FIG. 14 via the conversion circuit 89. However, this addition output is B
Since the data is picture data and is not used for generating a predicted image of another image, it is not stored in the frame memory 86.

After the B-picture image is output, the P-picture image data stored in the backward-predicted image portion 86b is read out, and the arithmetic unit 85 is passed through the motion compensation circuit 87.
Is supplied to. However, at this time, motion compensation is not performed.

The decoder 31 includes a prediction mode switching circuit 52 and a DC in the encoder 18 of FIG.
Although the circuits corresponding to the T mode switching circuit 55 are not shown, the processes corresponding to these circuits, that is, the configurations in which the signals of the odd field lines and the signals of the even fields are separated are originally mixed. The motion compensation circuit 87 executes the process of returning to the configuration described above as necessary.

Further, although the processing of the luminance signal has been described above, the processing of the color difference signal is similarly performed.
However, in this case, the motion vector used is one for the luminance signal, which is halved in the vertical and horizontal directions.

[0079]

By the way, when performing quantization and dequantization in conventional image signal coding, a value representing the fineness of quantization, that is, a quantization width (quantization step size). To use. 2 to 62 for this quantization width
Even numbers up to are used. In addition, there is a quantization characteristic (QUANT) as a value representing this quantization width. An integer value from 1 to 31 is used for this quantization characteristic, and the quantization width is a value obtained by doubling this quantization characteristic.

The above quantization width is a value necessary for compressing a general image into a target data amount. However, it is possible to compress an image whose statistical properties greatly deviate from the general image by using the above quantization width, for example, an image with extremely low pixel correlation or an image close to white noise in the frequency domain. It is very difficult to compress by DCT coding that utilizes the concentration of the coefficients of. That is, in this case, the image cannot be compressed to the target size even if the maximum value 31 of the quantization characteristic is used.

Further, when an image of very high quality is to be obtained, for example, when an image with almost no distortion (called lossless) is to be obtained, the quantization characteristic used is as large as 1, which is the minimum value. Since it is too large, it is not possible to perform the quantization capable of accurately restoring the image.

Further, in encoding the image signal, the compressed bit stream is often controlled to a target transmission rate. At this time, in a general linear quantizer, the above-mentioned quantization characteristic and the generated bit amount of image data quantized based on the above-mentioned quantization characteristic have a relation close to inverse proportion (more accurately, It is close to the logarithm).

Therefore, if the quantization characteristic is in a small range and the quantization characteristic is changed by 1, the generated bit amount changes greatly. For example, change the quantization characteristic from 1 to 2
When it is changed to, the generated bit amount becomes about half.
This indicates that when the quantization characteristic is in a small range, the quantization characteristic interval is too wide and it is difficult to finely control the generated bit amount.

On the contrary, when the quantization characteristic is in a large range, even if the quantization characteristic is changed by 1, the generated bit amount hardly changes. For example, when the quantization characteristic is changed from 30 to 31, the generated bit amount when the quantization characteristic is 31 changes by 5% of the generated bit amount when the quantization characteristic is 30. There is nothing to do. This indicates that the quantization characteristic interval is unnecessarily narrow when the quantization characteristic is in a large range.

In order to overcome the above problem, there is a method of mapping a number sequence of 1 to 31 which is simply increased as a quantization characteristic, to a non-linear number sequence instead of using it as it is. Table 7 shows the relationship between the quantization information for obtaining the above-mentioned quantization characteristics when the quantization characteristics are mapped to a non-linear sequence and the quantization characteristics based on this quantization information. The quantized information is called an index and may be represented by a numerical value.

[0086]

[Table 7]

By using the quantization characteristic of the non-linear sequence, it is possible to solve the problem caused by the range of the quantization characteristic as described above. However, since the quantization characteristic is the value of the table (mapping) converted into the non-linear sequence, the encoding device and the decoding device need a configuration for storing the value of the table. Therefore, the amount of hardware configuring the encoding device and the decoding device increases.

FIG. 20 shows a schematic structure of an inverse quantizer when the conventional quantization characteristic of a non-linear sequence is used.
ROM for storing the quantization characteristics of this non-linear sequence
A table unit 200 using (read only memory) or the like is prepared, and 8-bit quantization characteristics are read from the table unit 200. The multiplier 201 multiplies the quantization characteristic by an n-bit transform coefficient for the quantized image data, and dequantizes the quantized data. In the inverse quantizer, a large-scale circuit is required for the table section 200, and the multiplier 201 is also required to have a large-scale circuit.

That is, in the conventional image signal coding apparatus and decoding apparatus, since the values 1 to 31 or the values of the non-linear sequence shown in Table 7 are used as the quantization characteristics, A multiplier is required for the quantizer in the signal encoding device and the dequantizer in the image signal decoding device. This multiplier has a large circuit scale, which is a major obstacle to the configurations of the image signal encoding device and the image decoding device.

In view of the above situation, the present invention provides a code for an image signal which can use appropriate quantization characteristics when quantizing and dequantizing image data without increasing the scale of the constituent circuits. It is an object of the present invention to provide an encoding method and an encoding device, and an image signal decoding method and an encoding device.

[0091]

In the image signal coding method according to the present invention, at the time of quantization, 2 of the quantization information are used.
The first quantization information is a value for expressing the exponent of the power of 2 and the second is a value corresponding to a coefficient by which the power of 2 is multiplied.
The above-mentioned problem is solved by performing quantization based on a non-linear quantization characteristic (QUANT) represented by using a multiplication value of a power of 2 and the coefficient as the quantization information of 1.

Further, in the image signal decoding method according to the present invention, the value for expressing the exponent of power of 2 in the quantized information is set to the first quantized information during the inverse quantization. The value corresponding to the coefficient by which the power of 2 value is multiplied is used as the second quantization information, and the non-linear quantization characteristic is reproduced by the multiplication of the power value of 2 and the coefficient, and the encoded data is reproduced. The above-mentioned problem is solved by performing inverse quantization based on a non-linear quantization characteristic (QUANT).

In such an image signal coding method or decoding method as described above, k, 2 is added to the first quantized information which is a value for expressing the exponent of a power of 2 in the quantized information.
To obtain the quantization characteristic by using (i / 2 + j) for the second quantization information that is a value corresponding to the coefficient by which the power value of is multiplied, and set a constant multiple of this quantization characteristic as the quantization width. Is preferred.

Here, j and k are positive integers, and i is 0.
Alternatively, the value is represented by 1.

Further, the quantization characteristic QUANT is expressed by the equation: QUANT = (i / 2 + j) × 2 ^k +2 ^{(k + 2)} -4.

Further, the quantized information composed of the first quantized information and the second quantized information is expressed by a 5-bit code, and the quantized information k, j, i and the quantized characteristic are The relationship of is shown in Table 1 or Table 2 or Table 3 or Table 4.

Moreover, when the quantization characteristic is expressed by a binary number, effective bits are present in continuous 4 bits or 5 bits.

At this time, when the encoded data is inversely quantized, the encoded data is added three times, and the addition result is shifted by the bit determined by the first quantization information k.

Further, the first quantized information, which is a value for expressing the exponent of 2 in the quantized information, is a value corresponding to a coefficient by which the power value of 2 is multiplied by m. Quantization characteristics are obtained by using α _i for the quantization information of
It is preferable that the quantization width be a constant multiple of this quantization characteristic. Here, m is a power value (integer) required to express a desired quantization characteristic, and α _i (i = 1 to n) is a value represented by 0 or 1.

At this time, the quantization characteristic QUANT is QUANT = 2 ^(m-1) + α ₁ × 2 ^(m-2) + α ₂ × 2 ^(m-3) +
.. + α _n × 2 ^(mn-1) Here, n is a predetermined integer value that represents the accuracy of the quantization characteristic.

Further, when the number of bits required to represent the range that the first quantized information m can take is L, the quantized information including the first quantized information and the second quantized information. The coded information is represented by a code of (L + n) bits, and when the coded data is inversely quantized, the coded data is added n times and the addition result is shifted by L bits.

The quantized information consisting of the first quantized information and the second quantized information is represented by a 5-bit code, and the relationship between the quantized information m and α _i and the quantized characteristic. Are shown in Table 3 or Table 6.

At this time, when the encoded data is inversely quantized, the encoded data is added twice, and the addition result is shifted by 3 bits.

In the image signal coding method according to the present invention,
The amount of generated bits at the time of linear quantization is evaluated, and a linear / non-linear quantization switching signal indicating a quantization method is generated based on the evaluation result, and the linear / non-linear quantization switching signal indicates non-linear quantization. In this case, the value for expressing the exponent of 2 in the quantization information is the first quantization information, and the value corresponding to the coefficient by which the power of 2 is multiplied is used as the second quantization information. Quantization is performed based on a non-linear quantization characteristic (QUANT) represented by using a multiplication value of a power of 2 and the coefficient.

Here, the evaluation of the generated bit amount is characterized in that it is carried out in frame units.

Further, in the image signal coding apparatus according to the present invention, the coding unit for coding the input image signal by using the predetermined prediction image signal and the predetermined signal for the signal coded by the coding unit. A transform unit that performs a transform operation, and a coefficient that multiplies the signal from the transform unit by a value for expressing an exponent of a power of 2 in the quantized information, to the first quantized information and a power of 2 And a quantization unit that performs quantization based on a non-linear quantization characteristic (QUANT) represented by using a multiplication value of a power of 2 and the coefficient as a second quantization information value corresponding to The problem described above is solved by including a variable length coding unit that performs variable length coding on the quantized signal.

Further, in the image signal coding apparatus according to the present invention, the coding unit for coding the input image signal by using the predetermined prediction image signal and the predetermined signal for the signal coded by the coding unit. A conversion unit that performs a conversion operation, an evaluation unit that evaluates the amount of generated bits when linearly quantized, and a switch that generates a linear / non-linear quantization switching signal that indicates a quantization method based on the evaluation result of the evaluation unit. When the linear / non-linear quantization switching signal from the switching signal generating unit and the switching signal generating unit indicates linear quantization, the first quantizing unit that performs linear quantization on the signal from the converting unit; When the linear / non-linear quantization switching signal from the switching signal generation unit indicates non-linear quantization, the signal from the conversion unit is provided with a value for expressing an exponent of power of 2 in the quantization information. Quantized information of 1 is multiplied by a power of 2 A second quantum that is quantized based on a non-linear quantization characteristic (QUANT) represented by a product of a power of 2 and the above coefficient, with a value corresponding to the number as the second quantization information. And a variable length coding unit for variable length coding the signal quantized by the first quantizing unit or the second quantizing unit.

Here, the evaluation unit is characterized in that the generated bit amount is evaluated in frame units.

In the image signal decoding method according to the present invention,
In the case where the linear / non-linear quantization switching signal, which indicates which one of the linear quantization and the non-linear quantization should be performed, indicates the non-linear quantization, an exponent of 2 in the quantization information is set. A value corresponding to the first quantization information, a value corresponding to a coefficient by which a power of 2 is multiplied is used as second quantization information, and a quantization characteristic is obtained by multiplying the power of 2 by the coefficient. The above-mentioned problem is solved by reproducing and quantizing the quantized data based on the reproduced quantized characteristic (QUANT).

Here, the linear / non-linear quantization switching signal is switched on a frame-by-frame basis, whereby linear inverse quantization and nonlinear inverse quantization are performed on a frame-by-frame basis.

Further, in the image signal decoding apparatus according to the present invention, the variable length decoding unit for variable length decoding the transmitted image data and the exponent of 2 in the quantization information are expressed. Is used as the first quantization information, and the value corresponding to the coefficient by which the power of 2 is multiplied is used as the second quantization information.
A non-linear quantization characteristic (QUANT) is reproduced by multiplying the exponentiation value and the coefficient, and the quantized data from the variable length decoding unit is inversely quantized based on the reproduced non-linear quantization characteristic. The above-described problem is solved by having an inverse quantization unit for converting the data into a conversion unit and a conversion unit for performing a predetermined operation on the dequantized data.

Here, the dequantization section includes a table section for converting the first quantization information, and a shift means for shifting the second quantization information based on the first quantization information. And an addition unit for adding the output from the table unit and the output from the shift unit, and a multiplication unit for multiplying the quantized data by the output of the addition unit. To do.

Further, the inverse quantization section includes a table section for converting the first quantization information, an addition means for adding the output from the table section and the second quantization information,
It has a multiplication means for multiplying the output from the addition means and the quantized data, and a shift means for shifting the output from the multiplication means by a bit determined by the first quantization information. The multiplication means is composed of three stages of adders.

Further, the dequantization section includes a table section for converting the first quantization information, an addition means for adding the output from the table section and the second quantization information, and the variable section. Selecting means for switching and selecting one of the output from the adding means and the linear quantization information according to the linear / non-linear quantization switching signal decoded by the long decoding unit and transmitted together with the image data. Means for multiplying the output from the above and the quantized data, and only when the linear / non-linear quantization switching signal indicates non-linear quantization, the output from the above-mentioned multiplication means is output by the first quantization information. And a shift means for shifting by a defined bit, wherein the multiplying means comprises three stages of adders.

[0115]

The quantizing characteristic is converted into the value of the non-linear sequence, and the conversion method to the value of the non-linear sequence is appropriately selected.

At this time, the quantization and the dequantization are performed only by the multiplier having a small number of stages and the shift operation.

Quantization and dequantization are performed only by addition and shift operations.

[0118]

The preferred embodiments of the present invention will be described below with reference to the tables and the drawings.

In the first embodiment of the present invention, the value for expressing the exponent of power of 2 in the quantized information corresponds to the coefficient of the first quantized information and the power of 2 multiplied. The value of the second quantization information is used as the second quantization information, and the quantization characteristic (QUA) of the non-linear sequence represented by the multiplication value of the power of 2 and the above coefficient is used.
Quantization or dequantization based on NT). Therefore, k is added to the first quantized information and (i is added to the second quantized information).
When / 2 + j) is used, the quantization characteristic QUANT can be obtained from the following equation (1). Using this quantization characteristic, the image signal encoding device performs encoding, and the image signal decoding device performs decoding.

QUANT = (i / 2 + j) × 2 ^k + α (1) (α = 2 ^{(k + 2)} -4) j and k in the above formula (1) are positive integers, i Is a value represented by 0 or 1.

Here, since the variable length coded code is included in the bit stream coded by the MPEG system, all the variable length coded codes that may occur are generated when decoding is performed from the middle. However, a special code that can be uniquely decrypted is required. In the above bit stream, a code in which 23 or more 0s are consecutive is a special code. Therefore, in order to prevent 23 or more 0s from continuing in any combination of other variable-length coded codes, the quantization characteristic is set to the quantization information of the quantization information in order to limit the other variable-length coded codes. It is prohibited that all bits are 0 at the same time.

The sequence represented by the equation (1) is an arithmetic sequence having a power of 2 as a tolerance when the quantization information k is fixed, and the number represented by the quantization information j is p.
Then, the tolerance is switched in units of 2 × p.

Next, Table 8 shows the quantization characteristics that can be expressed by the equation (1).

[0124]

[Table 8]

In Table 8, as the quantization information, k is 2 bits, j is 2 bits, and i is 1 bit, which is a total of 5 bits, and the quantization characteristic corresponding to the quantization information is a decimal number. Numerical values and binary values are displayed. The 5 bits of the above quantization information are set to (Q1 Q2 Q3 Q4 Q5) from the most significant bit, and the first 2 bits of quantization information k (Q1 Q2) express the exponent of 2 in the equation (1). The first quantized information, which is the value of, the remaining 2-bit quantized information j (Q3 Q4) and 1-bit quantized information i (Q5) are multiplied by the power of 2 in the equation (1). The second quantized information is a value corresponding to the coefficient.

Further, the group of 8 units of quantization characteristics represented by X forms an arithmetic progression, and the tolerance is changed when the group of 8 units is switched. When the non-linear sequence shown in Table 8 is used, the conversion of the quantization information can be uniquely determined by the expression (1). Therefore, by expressing the quantization width with a value that can be expressed by the expression (1), A memory or the like for storing the conversion table becomes unnecessary.

Since the 5 bits indicated by the quantization information k, j, and i are the number of bits required to send the values 1 to 31 of the quantization characteristic, the conventional method for determining the quantization characteristic and the The compatibility can be maintained in the image signal encoding apparatus and the image signal decoding apparatus using the method for obtaining the quantization characteristic according to the present invention.

Further, in Table 8, the group of quantization characteristics is represented as X as described above. If the initial value of each group is α, the quantization characteristic is α by α and X.
It can be expressed as + X × 2 ⁿ (n is a natural number). As a result, the group of 8 units is 0 + X, 4+ from the beginning.
It is represented by 2X, 12 + 4X, 28 + 8X.

Next, FIG. 1 shows a schematic configuration of a quantization characteristic reproducing circuit in the image signal decoding apparatus for reproducing the quantization characteristic by converting into a non-linear sequence as shown in Table 8.

The sequence of X is the quantization information (Q3 Q4 Q
Input the value of 5) to the shifter 110, and quantize information (Q1 Q2)
By shifting using the value of, X × 2 ⁿ (n is a natural number) can be expressed. That is, the quantization information
If (Q1 Q2) is (0 0) 0 bits, if it is (0 1) 1 bit, if it is (1 0) 2 bits, (1
If 1), 3 bits worth of the above quantization information (Q3 Q4 Q
5) is shifted.

Here, in Table 9, the value of the quantization information (Q1 Q2) input to the table unit 111, the value S1 converted and output by the table unit 111, and the table unit as described above. The relationship with the value S2 sequentially read in 111 is shown.

[0132]

[Table 9]

Further, FIG. 2 shows the shifter 110 in FIG.
2 shows a schematic configuration of. AND gates 121 to 132
The quantized information (Q3 Q4 Q5) input to the switch is switched according to the shift amount generated based on the quantized information (Q1 Q2) in the shift amount generation unit 120, and further, the OR gates 133 and 136. , _EX. OR gates 134 and 135 are used to transfer the quantized information to the bit 0 output terminal b [0].
~ Bit 5 is output to the output terminal b [5].

Further, the values 0, 4, 12, 28 which are sequentially read out are stored in the table unit 111 in accordance with the quantization information (Q1 Q2).
The upper 3 bits of the read value and the upper 3 bits of the above X × 2 ⁿ (n is a natural number) are prepared in an adder 11
Add 2 Further, the 7-bit quantization characteristic QUANT can be reproduced by the 4 bits of the added value and the lower 3 bits of the value X × 2 ⁿ .

The constituent circuit for obtaining the nonlinear quantization characteristic as described above is much smaller than the conventional constituent circuit for obtaining the nonlinear quantization characteristic shown in Table 7. This is because the conventional non-linear quantization characteristic has no regularity, and a large number of gates are required by referring to all the tables.

Further, inverse quantization is performed by multiplying the quantization characteristic thus obtained by the conversion coefficient as the quantized image data (quantized data). In this case, as is apparent from Table 8 above, since 4 bits out of 7 bits are effective bits, at the time of multiplication of this quantization characteristic and the transform coefficient as the quantized data (quantized data) of the image signal, Multiplication can be performed using a three-stage adder.

Next, FIG. 3 shows a schematic configuration of an inverse quantization circuit in the image signal encoding apparatus and the image decoding apparatus. The quantization information (Q1 Q2) input to the table unit 141 is shown in Table 1.
The value is converted to a value S3 indicated by 0 and sent to the adder 140.

[0138]

[Table 10]

In the adder 140, the value S3 and the quantized information (Q3 Q4 Q5) are added, and the added value is sent to the signal switch 144. In addition, the signal switching unit 144 is provided with quantization information (Q1
Q2 Q3 Q4 Q5) and 0, which is the most significant bit of quantization information when performing non-linear quantization, are input. Further, the signal switching unit 144 is input with a linear / non-linear quantization switching signal for selecting whether to perform linear quantization or non-linear quantization.

However, the linear quantization mentioned here means that the correspondence between the value expressed as a binary value of the quantization information and the quantization width (step size) is linear, and the non-linear quantization is , That the correspondence between the value represented as a binary value of the quantization information and the quantization width is non-linear.

In this signal switch 144, when linear quantization is selected by the linear / non-linear quantization switching signal, the quantization information (Q1 Q2 Q3
Q4 Q5) is selected and sent to the multiplier 142. But,
When the non-linear quantization is selected by the linear / non-linear quantization switching signal, the 4-bit quantization information from the adder 140 and 0 which is the most significant bit of the quantization information are selected and the multiplier 142 is selected. Sent. The multiplier 142 multiplies the input quantized information by an n-bit transform coefficient, and outputs the multiplied value to the shifter 143.

The shifter 143 has a quantization information (Q1
Q2) and the linear / non-linear quantization switching signal are input. Therefore, in the shifter 143, when the linear quantization is selected by the linear / non-linear quantization switching signal, the output from the multiplier 142 is directly output as the reproduction data. However, when the non-linear quantization is selected by the linear / non-linear quantization switching signal, the multiplier 142 has the shift amount shown in Table 11 obtained based on the inputted quantization information (Q1 Q2). Output is shifted and output as reproduction data.

[0143]

[Table 11]

In Table 11, the value of the quantization information (Q1 Q2) when the linear quantization is selected by the linear / non-linear quantization switching signal is related to the shift amount in any combination. Indicates not to.

A is the reproduction value of the data when the quantized data is inversely quantized, and Co is the conversion coefficient as the quantized data.
When the eff and the quantization width are SP, the reproduction value A can be expressed by the following equation (2). A = Coeff × SP = Coeff × (2 × QUANT) (2)

Here, when the non-linear quantization is selected, the equation (2) for obtaining the above-mentioned quantization characteristic QUANT is modified as follows. QUANT = ((i / 2 + j) + (4-4 / 2 k)) × 2 k ··· (3) The (3) where the (i / 2 + j) is a second input to the adder 140 Corresponding to the quantization information (Q3 Q4 Q5), (4-4 / 2 ^k ) corresponds to the output from the table unit 141, and 2 ^k indicates the shift amount in the shifter 143. There is. Therefore, the reproduction value A is expressed by the following equation (4).

A = Coeff × ((i / 2 + j) + (4-4 / 2 ^k )) × 2 ^{(k + 1)} ·· (4)

The structure of the shifter 143 for obtaining the reproduction value A is simple, and the multiplier 142 uses n bits of the conversion coefficient and 4 bits and the most significant bit of the output data from the switch 144. A relatively low number of stages that can be multiplied by is sufficient.

Here, in the above-described first embodiment, the maximum value that the quantization characteristic can take is 56.0, and the quantization width is 112. However, in an actual image, for example, when white noise is input, a larger quantization characteristic is required. For this,
As a second embodiment, it can be dealt with by two methods described later.

In the first method, the quantization information is "0000".
When it is 0 ”(binary representation), this quantization information is currently unused, so when this quantization information is“ 00000 ”, values such as 64, 96, and 128 are assigned to the quantization characteristic and processing is performed. When the quantization characteristic is assigned to 64 or 128, the processing is easy because the shifter only needs to perform the shift in the multiplication in the inverse quantization, and the quantization characteristic is assigned to 96. In this case, the number of adders is only one, and the processing is similarly easy.

In the second method, the quantization information is "0000".
When using 0 "(binary number representation), a long zero sequence may occur, so the quantization information is" 11111 ".
At this time, a value such as 64, 96, or 128 is assigned to the quantization characteristic and the processing is performed.

Furthermore, as a third embodiment, 28 + 8X
Table 12 shows the quantization characteristics in the case where the shift amount of the fourth group indicated by is largely changed.

[0153]

[Table 12]

In the quantized information of the fourth group indicated by 28 + 16X in Table 12, the maximum value of the shift amount indicated by the quantized information k is intentionally increased by ignoring the continuity of the quantized information k. Required by setting. As a result, the maximum value that the quantization characteristic can take is 84.0. This value is a value that can sufficiently deal with a special input such as white noise. This method is also a preferable method for controlling the encoding because it is possible to prepare continuous quantization characteristics up to the maximum value. Also in the third embodiment, as in the first embodiment, the quantization information is prohibited from being "00000".

Next, FIG. 4 shows a schematic structure of a quantization characteristic reproducing circuit in the image signal decoding apparatus which reproduces the quantization characteristic by converting into a non-linear sequence as shown in Table 12.

The quantized information (Q3 Q4 Q5) is transferred to the shifter 15
It is possible to express X × 2 ⁿ (n is a natural number) by inputting 0 and shifting using the value of the quantization information (Q1 Q2). That is, if the quantized information (Q1 Q2) is (0 0), it is 0 bit, if it is (0 1), it is 1 bit,
2 bits for (1 0), 4 for (1 1)
The quantization information (Q3 Q4 Q5) is shifted by a bit.

FIG. 5 shows a schematic structure of the shifter 150 shown in FIG. The quantization information (Q3 Q4 Q5) input to each AND gate 161 to 172 is the shift amount generation unit 1
At 60, switching is performed according to the shift amount generated based on the quantization information (Q1 Q2), and the OR gate 17
3,175,176, by the intervention of the E _X. OR gate 174, the quantization information is outputted to the bit 0 output terminal b [0] ~ bit 6 output terminal b [6].

Values 0, 4, 12, and 28 to be sequentially read out are prepared in the table 151 in accordance with the quantized information (Q1 Q2), and the upper 4 bits of the read out value and the above X × 2.
^The adder 152 adds the upper 4 bits of ⁿ (n is a natural number). Furthermore, by using 5 bits of the added value and the lower 3 bits of the value X × 2 ⁿ , a quantization characteristic of 8 bits is obtained.
QUANT can be played. The table section 151
The values sequentially read from are the values shown in Table 9 similar to the first embodiment.

Next, FIG. 6 shows a schematic configuration of an inverse quantization circuit in the image signal encoding device and the image signal decoding device. The quantization information (Q1 Q2) input to the table unit 181 is shown in Table 1
3 is converted into a value S4 and sent to the adder 180.

[0160]

[Table 13]

In the adder 180, the table unit 1
The value S4 from 81 and the quantized information (Q3 Q4 Q5) are added and multiplied by the n-bit conversion coefficient in the multiplier 182. The multiplied value is shifted by the shifter 183 based on the quantized information (Q1 Q2) to reproduce the image data. The number of bits generated from the multiplier 182 of the inverse quantization circuit in the third embodiment is different from the number of bits generated from the multiplier 142 of the inverse quantization circuit in the first embodiment described above.

Therefore, the reproduction value A of the image data when the quantized information k in the third embodiment has a value of 0, 1, or 2, the quantized information k can be obtained by the above equation (4). When A has a value of 3, the reproduction value A of the image data is
It is obtained from the equation (5). A = Coeff × ((i / 2 + j) +1.75) × 2 ⁵ (5)

The structure of the shifter 183 for obtaining the reproduction value A is simple, and the multiplier 182 multiplies n bits of the conversion coefficient by 5 bits of the output data from the adder 180. It is sufficient that the number of steps is relatively low.

Next, a fourth embodiment will be described. In the image signal encoding method of the present invention, the image signal is quantized by the quantization characteristic QUANT represented by the following equation (6).

QUANT = 2 ^(m-1) + α ₁ × 2 ^(m-2) + α ₂
× 2 ^(m-3) + ·· + α _n × 2 ^(mn-1) (6) Here, m is a numerical value (integer) necessary to represent a desired quantization characteristic, α _i ( i = 1 to n) is a value represented by 0 or 1, and n is a predetermined integer value representing the accuracy of the quantization characteristic.

In the fourth embodiment, as a concrete example of the expression (6), the power value m is an integer from 0 to 7,
The precision n of the quantization characteristic is a value represented by 2, and the signal is quantized. In the above formula (6), (m-1) is used instead of the power value m, but they are essentially the same.

Table 14 shows the quantization characteristics that can be expressed by the equation (6).

[0168]

[Table 14]

As described above, when the non-linear sequence of numbers shown in Table 14 is used, since the mapping can be uniquely determined by the expression (6), a memory for storing the mapping becomes unnecessary.

Further, in the case of using the mapping to the non-linear sequence of Table 14, the following codes are transmitted to send the quantization information in the equation (6). First, since it is necessary to send an integer from 0 to 7 in order to send the quantized information m, 3 bits are required. Furthermore, in order to send the quantized information α ₁ and α ₂ , one bit is required for each, so that a total of 5 bits are required. These 5 bits are 1
Since the number of bits required to send the values from 1 to 31 is just correct, compatibility is achieved in a system that employs both the conventional quantization characteristic and the quantization characteristic of the present invention.

Here, the 5-bit quantization information (Q1 Q
An example of the configuration of 2 Q3 Q4 Q5) is shown. First 3 bits (Q1
Q2 Q3) represents the quantized information m using a binary number, and then bit by bit (Q4 Q5) represents the quantized information α ₁ and α ₂ .

Q1 Q2 Q3 Q4 Q5: 5-bit Q1 Q2 Q3: 000-111: Quantization information m Q4: 0 or 1: Quantization information α ₁ Q5: 0 or 1: Quantization information α ₂

Next, consider the inverse quantization in the case of using the mapping to the non-linear sequence of Table 14. When the non-linear sequence of Table 14 is expanded to a binary number, the number of bits in which 1 is set is 3
There is only one. Therefore, the maximum number of additions is two, and the number of adders (adders) can be two. Further, since the bits in which 1 is set are not random but are always continuous, it is sufficient to shift to a desired position by the shifter after the two-stage adder (adder). Such an inverse quantization circuit of the present invention is shown in FIG. The dequantization device of the present invention is composed of two adders (full adders) 190 and 191 and a shifter 192. In this shifter 192, (Q1 Q2 Q3)
The value is shifted to the left by the number of bits shown in Table 15. At this time, 0 is filled in the LSB.

A specific example is shown here. For example, when the quantization width is 20 and the DCT coefficient is 100, this DCT
The case of quantizing the coefficient will be described. At this time, the quantization characteristic becomes 5, and if this value is encoded with 9 bits, it becomes "0".
It is encoded to 00000101 "and transmitted.
When the quantization width 20 is coded into 5 bits by the method of the present invention, 20 = 16 + 4, 16 = 2 ⁴ , 4 = 2 ^2, so m
= 5, α1 = 0, α2 = 1, and the code of 5 bits is “10101”.

On the decoding side, the quantized value (quantized data) "000000101" and the quantization width "1010".
When 1 "is received, these are input to the inverse quantization circuit. That is, in FIG. 7, a0 ... a8 =" 0000 ".
0101 "and Q1 Q2 Q3 =" 101 ": quantized information m Q4 =" 0 "Q5 =" 1 ".

At this time, in the inverse quantization circuit of FIG.
Since "4" is "1", a0 ... a8 in the uppermost stage is input to the adder 190 as it is, but since Q4 = "0", 0 is output in a0 ... a8 in the next stage. Not entered in 190. Further, a0 ... a8 in the third row is added to this addition result. Therefore, the addition result is "000
0000111001 ". This value corresponds to the shifter 19
2 is input and is shifted by 2 bits according to Q1 Q2 Q3 = "101", and the output from the shifter 192 is "0000".
01100100 ", and thus the quantized value" 10 "
0 "is obtained. The shift amount in the shifter 192 is shown in Table 15, and the calculation method of the shift amount in this specific example is shown in Table 16.

[0177]

[Table 15]

[0178]

[Table 16]

Next, a fifth embodiment will be described. In the fourth embodiment, as shown in Table 14, the quantization characteristic of the precision up to the third decimal place of the binary number is defined, but the precision of the quantization characteristic accepted by the quantization circuit is independent. Determined. For example, as an example, when the precision of the quantization characteristic received by the quantization circuit is up to the first decimal place in the binary number display, the quantization information indicating the quantization characteristic of which the precision is too high among the quantization characteristics in Table 14 is shown. Prohibit Table 17 shows the quantization characteristics thus limited. In this fifth embodiment,
It is not possible to use prohibited quantized information.

[0180]

[Table 17]

Further, instead of prohibiting the quantized information showing the quantized characteristic with too high accuracy, the quantized information is
The case where the assignment is changed so as to represent the quantization characteristic of the nearest acceptable precision is the sixth embodiment. This is shown in Table 18. In Table 18, for example, the quantization information indicating the quantization characteristic of 0.5 is “000xx”. Where x
Represents don't care and indicates that the bit at that position may be 0 or 1. In the sixth embodiment, there is no prohibited quantization information as in the fifth embodiment.

[0182]

[Table 18]

Since the MPEG-coded bit stream includes a variable length coding code (VLC), when decoding from the middle, it is surrounded by all possible variable length coding codes. Even so, a special code that can be uniquely decrypted is required. In an MPEG encoded bitstream, a code that is followed by 23 or more 0s is a special code, so that any combination of other variable length coded codes is followed by 23 or more 0s. As such, other variable length coded codes are restricted.

Therefore, the quantized information composed of only 0 is not used as much as possible. Therefore, Tables 14, 17, 1
The quantized information "00000" in 8 becomes a problem. To solve this, for example, in Table 14, "0" and "
By inverting 1 ”, the large quantization characteristic“ 112 ”which is considered to be rarely used is prohibited. This example is shown in Table 19. Also, the precision of Table 18 is doubled to quantize information. Table 20 shows the case of inverting.

[0185]

[Table 19]

[0186]

[Table 20]

Next, a coding device and a decoding device for an image signal when switching between linear quantization and non-linear quantization when using quantization information will be described.

The schematic configuration of the image signal coding apparatus according to the present invention is the same as that of the conventional coding apparatus shown in FIG. 16, but the dequantization circuit 60 is shown in FIGS. 7
The quantization circuit 57 has any one of the schematic configurations shown in FIG. 8 and the quantization circuit 57 has the schematic configuration shown in FIG.

The quantizing circuit 57 shown in FIG.
The signal from the T circuit 56 is input to the maximum coefficient selection circuit 2
Sent to 10. In the maximum coefficient selection circuit 210, the value obtained by dividing the maximum coefficient by the maximum quantization width 62 used in linear quantization is compared with the maximum value 256 of the quantization level, and the obtained value is the maximum quantization. If the value is equal to the width or larger than the maximum quantization width, the quantization characteristic selection circuit 211 selects the quantization characteristic for linear quantization. At this time, a quantization selection signal indicating that linear quantization has been selected from the quantization characteristic selection circuit 211 is a signal changeover switch 212 and the variable length coding circuit 58 of FIG.
Is output to. Therefore, since the signal changeover switch 212 is changed over to the terminal a side, the quantization characteristic selection circuit 211
Is output to the linear quantization circuit 213 via the terminal a of the signal changeover switch 212. The linear quantization circuit 213 performs linear quantization by the linear quantization characteristic, and the linearly quantized data is output to the variable length coding circuit 58 and the inverse quantization circuit 60.

Since the quantization width (scale) is also input to the variable length coding circuit 58 together with the quantization characteristic selection signal, the variable length coding circuit 58 uses the quantization width to change the quantization width. Long encoding is performed.

On the other hand, the maximum coefficient selection circuit 21
In the comparison between the obtained value at 0 and the maximum value of the quantization level, if the obtained value is a value smaller than the maximum quantization width, the quantization characteristic selection circuit 211 uses the quantization for nonlinear quantization. A characterization characteristic is selected. Therefore, the quantization selection signal indicating that the nonlinear quantization is selected from the quantization characteristic selection circuit 211 is output to the signal changeover switch 212, and the signal changeover switch 212 is changed over to the terminal b side. The output from the quantization characteristic selection circuit 211 is output to the nonlinear quantization circuit 214 via the terminal b of the signal changeover switch 212, and nonlinear quantization is performed by the nonlinear quantization characteristic. The non-linear quantized data is output to the variable length coding circuit 58 and the dequantization circuit 60.

Further, similarly to the case where the linear quantization is performed, since the quantization width is also input to the variable length coding circuit 58 together with the quantization characteristic selection signal, the variable length coding circuit 58 Variable length coding is performed using the quantization width.

Here, since the linear / non-linear quantization switching signal is switched in frame units, the linear quantization and non-linear quantization are performed in frame units.

In the maximum coefficient selection circuit 210, the MPE
It is also possible to determine whether to perform linear quantization or non-linear quantization by using a flag indicating the quantization characteristic to be used, which is defined by the G method.

Further, the maximum coefficient selection circuit 210 can also determine whether to perform linear quantization or non-linear quantization using the dynamic range.

The schematic configuration of the image signal decoding apparatus according to the present invention is similar to that of the conventional decoding apparatus shown in FIG. 19, but the inverse quantization circuit 83 is shown in FIGS. , Has any one of the schematic configurations shown in FIG.

The above-mentioned embodiment is an example of the present invention, and it goes without saying that various other configurations can be adopted without departing from the gist of the present invention.

[0198]

As is apparent from the above description, in the image signal encoding apparatus according to the present invention, an encoding unit that encodes an image signal using a predetermined predicted image signal, and the encoding unit described above. A conversion unit that performs a predetermined conversion operation on the signal encoded by the above, an evaluation unit that evaluates the amount of generated bits when linearly quantized, and a linear / quantification method that indicates the quantization method based on the evaluation result of the evaluation unit. When the switching signal generation unit that generates a non-linear quantization switching signal and the linear / non-linear quantization switching signal from the switching signal generation unit indicate linear quantization, the signal from the conversion unit is subjected to linear quantization. When the linear / non-linear quantization switching signal from the first quantizing unit and the switching signal generating unit indicates non-linear quantization, the signal from the converting unit is a power of 2 of the quantization information. The first value to express the exponent
Of the non-linear quantization characteristic represented by using the multiplication value of the power value of 2 and the coefficient as the second quantization information, the value corresponding to the coefficient by which the power value of 2 is multiplied. QUA
NT) and a second quantization unit for performing quantization on the basis of (NT), and a variable length coding unit for variable length coding the signal quantized by the first quantization unit or the second quantization unit. By having it, the image data can be quantized or dequantized with a wide range of suitable and accurate quantization characteristics.

Further, in the image signal decoding apparatus according to the present invention, the variable length decoding unit for variable length decoding the transmitted image data and the exponent of 2 in the quantized information are expressed. Is used as the first quantization information, and the value corresponding to the coefficient by which the power of 2 is multiplied is used as the second quantization information.
A non-linear quantization characteristic (QUANT) is reproduced by multiplying the exponentiation value and the coefficient, and the quantized data from the variable length decoding unit is inversely quantized based on the reproduced non-linear quantization characteristic. Dequantizing image data with a quantization characteristic of appropriate precision over a wide range by including an inverse quantization unit that converts the image data and a conversion unit that performs a predetermined operation on the dequantized data. You can

Therefore, a sufficiently large quantization characteristic can be used when encoding an image requiring a large quantization characteristic, and a sufficiently small quantization characteristic can be used when decoding a high quality image. Can be used.

Further, it is possible to obtain an appropriate quantization characteristic for controlling the bit amount generated by encoding with high accuracy.

Further, since it is not necessary to store a non-linear sequence of numbers for the quantization characteristic, the scale of the constituent circuit does not increase, and the quantization and the dequantization are performed only by the multiplier having a small number of stages and the shift operation. Therefore, the scale of the multiplier can be reduced to half that of the conventional image signal encoding method, encoding apparatus, decoding method, and decoding apparatus.

[Brief description of drawings]

FIG. 1 is a diagram showing a schematic configuration of a quantization characteristic reproducing circuit in an image signal encoding apparatus and an image signal decoding apparatus according to the present invention.

2 is a diagram showing a schematic configuration of a shifter 110 in FIG.

FIG. 3 is a diagram showing a schematic configuration of an inverse quantization circuit in an image signal encoding apparatus and an image signal decoding apparatus according to the present invention.

FIG. 4 is a diagram showing a schematic configuration of a quantization characteristic reproducing circuit in an image signal encoding device and a decoding device of a third embodiment according to the present invention.

5 is a diagram showing a schematic configuration of a shifter 150 in FIG.

FIG. 6 is a diagram showing a schematic configuration of an inverse quantization circuit in an image signal encoding device and an image signal decoding device according to a third embodiment of the present invention.

FIG. 7 is a diagram showing a schematic configuration of an inverse quantization circuit in an image signal encoding device and a decoding device according to a fourth embodiment of the present invention.

FIG. 8 is a diagram showing a schematic configuration of a quantization circuit in an image signal encoding apparatus according to the present invention.

FIG. 9 is a diagram illustrating the principle of high efficiency encoding.

[Fig. 10] Fig. 10 is a diagram for describing picture types in the case of compressing image data.

FIG. 11 is a diagram illustrating the principle of encoding a moving image signal.

FIG. 12 is a diagram illustrating a GOP Structure of an image signal.

FIG. 13 is a diagram for explaining an input order, an encoding order, a decoding order, and an output order of image signals.

FIG. 14 is a block circuit diagram showing a configuration example of a conventional image signal encoding device and image signal decoding device.

15 is a diagram illustrating a format conversion operation of the format conversion circuit 17 of FIG.

16 is a block circuit diagram showing a configuration example of an encoder 18 of FIG.

17 is a diagram illustrating the operation of the prediction mode switching circuit 52 in FIG.

18 is a diagram for explaining the operation of the DCT mode switching circuit 55 in FIG.

19 is a block circuit diagram showing a configuration example of the decoder 31 of FIG.

FIG. 20 is a diagram showing a schematic configuration of a conventional non-linear quantization circuit.

[Explanation of symbols]

110, 143, 150, 183 ... Shifter 111, 141, 151, 181, ... Table unit 112, 140, 152, 180 ... Adder 142, 182 .... Multiplier 144: Signal switch 190, 191: Adder 192: ...・ ・ ・ Shifter 210 ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ Maximum coefficient selection circuit 211 ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ Quantization characteristic selection circuit 212 ・ ・ ・・ ・ ・ ・ ・ ・ ・ ・ ・ Signal changeover switch 213 ・ ・ ・ ・ ・ ・ ・ ・ Linear quantization circuit 214 ・ ・ ・ ・ ・ ・ ・ ・・ Nonlinear quantization circuit

Continuation of the front page (31) Priority claim number Japanese Patent Application No. 5-66550 (32) Priority date Hei 5 (1993) March 25 (33) Priority claim country Japan (JP)

Claims (40)

[Claims] 1. A method of encoding an image signal for quantizing and encoding an input image signal, wherein at the time of the quantization, a value for expressing an exponent of 2 in the quantization information is expressed as a first value. Non-linear quantization characteristic represented by using the multiplication value of the power value of 2 and the coefficient as the second quantization information, the value corresponding to the coefficient by which the power value of 2 is multiplied. An image signal encoding method characterized by performing quantization based on (QUANT). 2. A value corresponding to a coefficient by which the first quantized information, which is a value for expressing the exponent of 2 in the quantized information, is multiplied by a power value of k, 2. 2. The image signal coding method according to claim 1, wherein a quantization characteristic is obtained by using (i / 2 + j) for the quantization information of 1. and the quantization width is a constant multiple of the quantization characteristic. Here, j and k are positive integers, and i is a value represented by 0 or 1. 3. The image signal according to claim 2, wherein the quantization characteristic QUANT is expressed by an equation shown in QUANT = (i / 2 + j) × 2 ^k +2 ^{(k + 2)} −4. Encoding method. 4. The method of encoding an image signal according to claim 3, wherein the quantized information composed of the first quantized information and the second quantized information is represented by a 5-bit code. . 5. The method of encoding an image signal according to claim 4, wherein the relationship between the quantized information k, i, j and the quantized characteristic is expressed in Table 1 below. [Table 1] 6. The method of encoding an image signal according to claim 5, wherein when the quantization characteristic is expressed by a binary number, effective bits are present in 4 consecutive bits. 7. The method of encoding an image signal according to claim 4, wherein the relationship between the quantized information k, i, j and the quantized characteristic is expressed in Table 2 below. [Table 2] 8. The method of encoding an image signal according to claim 7, wherein when the quantization characteristic is represented by a binary number, effective bits are present in consecutive 5 bits. 9. A second value which is a value corresponding to a coefficient by which the first quantized information, which is a value for expressing an exponent of 2 of the quantized information, is multiplied by a power value of m, 2. coding method according to claim 1, the image signal, wherein the Determination quantization characteristics by using alpha _i to the quantization information, the constant multiple of the quantization characteristic and the quantization width. Here, m is a numerical value (integer) required to express a desired quantization characteristic, α
_i (i = 1 to n) is a value represented by 0 or 1. 10. The quantization characteristic QUANT is QUANT = 2 ^(m-1) + α ₁ × 2 ^(m-2) + α ₂ × 2 ^(m-3) +
The image signal encoding method according to claim 9, wherein the image signal is represented by the equation: + α _n × 2 ^(mn-1) . Here, n is a predetermined integer value that represents the accuracy of the quantization characteristic. 11. When the number of bits required to represent the range that the first quantized information m can take is L, the quantized information including the first quantized information and the second quantized information. 11. The image signal encoding method according to claim 10, wherein the encoded information is represented by a code of (L + n) bits. 12. The method of encoding an image signal according to claim 11, wherein the quantization information composed of the first quantization information and the second quantization information is represented by a 5-bit code. . 13. The method of encoding an image signal according to claim 12, wherein the relationship between the quantized information m and α _i and the quantized characteristic is expressed in Table 3 below. [Table 3] 14. An image signal is encoded using a predetermined predicted image signal, the signal obtained by subjecting the encoded signal to a predetermined operation is quantized, and the quantized signal is of variable length. In the encoding method of the image signal to be encoded, the amount of generated bits at the time of linear quantization is evaluated, and a linear / non-linear quantization switching signal indicating the quantization method is generated based on the evaluation result, and the linear / non-linear quantization switching signal is generated. When the non-linear quantization switching signal indicates non-linear quantization, the value for expressing the exponent of power of 2 in the quantization information is used as the first quantization information and the coefficient by which the power of 2 is multiplied. An image characterized by performing quantization based on a non-linear quantization characteristic (QUANT) represented by using a multiplication value of a power of 2 and the above coefficient, with a corresponding value as second quantization information. Signal coding method. 15. The method of encoding an image signal according to claim 14, wherein the evaluation of the generated bit amount is performed in frame units. 16. An image signal encoding apparatus for quantizing and encoding an input image signal, comprising: an encoding unit for encoding the input image signal using a predetermined predicted image signal; and an encoding unit for encoding by the encoding unit. A conversion unit for performing a predetermined conversion operation on the generated signal, and a value for expressing an exponent of a power of 2 in the quantization information in the signal from the conversion unit as the first quantization information and a power of 2 The value corresponding to the coefficient by which the value is multiplied is used as the second quantization information, and the quantization is performed based on the non-linear quantization characteristic (QUANT) represented by the product value of the power of 2 and the coefficient. An image signal coding apparatus, comprising: a quantizing unit for performing a variable length coding; and a variable length coding unit for variable length coding the quantized signal. 17. An image signal coding apparatus for quantizing and coding an input image signal, comprising: a coding section for coding the input image signal using a predetermined predicted image signal; and a coding section for coding by the coding section. A conversion unit that performs a predetermined conversion operation on the generated signal, an evaluation unit that evaluates the amount of generated bits when linearly quantized, and a linear / non-linear quantization that indicates a quantization method based on the evaluation result of the evaluation unit. A switching signal generation unit that generates a switching signal, and a linear / non-linear quantization switching signal from the switching signal generation unit that indicates linear quantization, the first signal that performs linear quantization on the signal from the conversion unit. When the linear / non-linear quantization switching signal from the quantizing unit and the switching signal generating unit indicates non-linear quantization, the signal from the converting unit expresses an exponent of 2 in the quantized information. The value for performing the first quantization information, Based on the non-linear quantization characteristic (QUANT) represented by using the value corresponding to the coefficient by which the power value of 2 is multiplied as the second quantization information, the multiplication value of the power value of 2 and the coefficient is used. A second quantization unit for performing quantization, and a variable length coding unit for variable length coding the signal quantized by the first quantization unit or the second quantization unit. An image signal encoding device characterized by: 18. The image signal coding apparatus according to claim 17, wherein the evaluation section evaluates the generated bit amount in frame units. 19. A decoding method of an image signal for dequantizing transmitted coded data, decoding the dequantized data to restore image data, wherein in the dequantization, A value for expressing an exponent of power of 2 in the quantized information is used as the first quantized information, and a value corresponding to a coefficient by which the value of the power of 2 is multiplied is used as the second quantized information. Decoding an image signal characterized by reproducing a non-linear quantization characteristic by multiplying a value by the coefficient and dequantizing encoded data based on the reproduced non-linear quantization characteristic (QUANT). Method. 20. A value corresponding to a coefficient by which the first quantized information, which is a value for expressing an exponent of 2 in the quantized information, is multiplied by a power value of k, 2 20. The method for decoding an image signal according to claim 19, wherein the quantization characteristic is obtained by using (i / 2 + j) for the quantization information of 1. and the quantization width is a constant multiple of the quantization characteristic.
Here, j and k are positive integers, and i is a value represented by 0 or 1. 21. The image signal according to claim 20, wherein the quantization characteristic QUANT is expressed by an equation represented by QUANT = (i / 2 + j) × 2 ^k +2 ^{(k + 2)} −4. Decryption method. 22. The image signal decoding method according to claim 21, wherein the quantized information composed of the first quantized information and the second quantized information is represented by a 5-bit code. . 23. When dequantizing the coded data, the coded data is added three times, and the addition result is shifted by a bit determined by the first quantization information k. Item 23. A method for decoding an image signal according to Item 22. 24. The method of decoding an image signal according to claim 22, wherein the relationship between the quantized information k, i, j and the quantized characteristic is expressed in Table 4 below. [Table 4] 25. The method of decoding an image signal according to claim 22, wherein the relationship between the quantized information k, i, j and the quantized characteristic is expressed in Table 5 below. [Table 5] 26. A second value corresponding to a coefficient by which the first quantized information, which is a value for expressing an exponent of 2 in the quantized information, is multiplied by a power value of m, 2. 20. The method of decoding an image signal according to claim 19, wherein a quantization characteristic is obtained by using α _i for the quantization information of, and a constant multiple of the quantization characteristic is used as the quantization width. Where m
Is a numerical value (integer) necessary to represent a desired quantization characteristic, and α _i (i = 1 to n) is a value represented by 0 or 1. 27. The quantization characteristic QUANT is QUANT = 2 ^(m-1) + α ₁ × 2 ^(m-2) + α ₂ × 2 ^(m-3) +
The image signal decoding method according to claim 26, characterized in that the image signal is represented by the equation shown in + α _n × 2 ^(mn-1) . Here, n is a predetermined integer value that represents the accuracy of the quantization characteristic. 28. When the number of bits required to represent a range that the first quantized information m can take is L, the quantized information including the first quantized information and the second quantized information. 28. The image signal decoding method according to claim 27, wherein the encoded information is represented by a code of (L + n) bits. 29. Decoding an image signal according to claim 28, wherein when the coded data is inversely quantized, the coded data is added n times and the addition result is shifted by L bits. Method. 30. Decoding of an image signal according to claim 28, wherein the quantized information composed of the first quantized information and the second quantized information is represented by a code of (3 + 2) bits. Method. 31. Decoding of an image signal according to claim 29, wherein when dequantizing the coded data, the coded data is added twice and the addition result is shifted by 3 bits. Method. 32. The method of decoding an image signal according to claim 30, wherein the relationship between the quantized information m and α _i and the quantized characteristic is expressed in Table 6 below. [Table 6] 33. An image signal decoding method for dequantizing data obtained by variable length decoding of transmitted image data, decoding the dequantized data, and restoring the image data, In the case where the linear / non-linear quantization switching signal, which indicates which one of the linear quantization and the non-linear quantization should be performed, indicates the non-linear quantization, an exponent of 2 in the quantization information is set. A value corresponding to the first quantization information, a value corresponding to a coefficient by which a power of 2 is multiplied is used as second quantization information, and a quantization characteristic is obtained by multiplying the power of 2 by the coefficient. A method for decoding an image signal, which is characterized in that the reproduced and quantized data is inversely quantized based on the reproduced quantized characteristic (QUANT). 34. The image signal according to claim 33, wherein linear inverse quantization and nonlinear inverse quantization are performed in frame units by switching the linear / nonlinear quantization switching signal in frame units. Decryption method. 35. Dequantizing the transmitted image data,
In the image signal decoding device for decoding the dequantized data to restore the image data, a variable length decoding unit for variable length decoding the transmitted image data, and 2 of the quantization information The first quantization information is used as a value for expressing the exponent of the power of 2 and the value corresponding to the coefficient by which the power of 2 is multiplied is used as the second quantization information. And a dequantization unit that dequantizes the quantized data from the variable length decoding unit based on the reconstructed nonlinear quantization property, An image signal decoding apparatus, comprising: a conversion unit that performs a predetermined operation on the inversely quantized data. 36. The dequantization unit, a table unit for converting the first quantization information, a shift unit for shifting the second quantization information based on the first quantization information, It is characterized by further comprising: adding means for adding the output from the table section and the output from the shift means; and multiplying means for multiplying the quantized data by the output of the adding means. The image signal decoding apparatus according to claim 35. 37. The dequantization unit, a table unit for converting the first quantization information, an addition unit for adding an output from the table unit and the second quantization information, and the addition Multiplying the output from the means with the quantized data, and the output from the multiplying means into the first
37. The image signal decoding apparatus according to claim 36, further comprising a shift means for shifting by a bit determined by the quantized information of. 38. The image signal decoding apparatus according to claim 37, wherein said multiplication means is composed of a three-stage adder. 39. The dequantization unit, a table unit for converting the first quantization information, an addition unit for adding an output from the table unit and the second quantization information, the variable Selection means for switching and selecting one of the output from the adding means and the linear quantization information according to the linear / non-linear quantization switching signal decoded by the long decoding unit and transmitted together with the image data, and the selecting means. Means for multiplying the output from the quantized data with the linear /
Only if the non-linear quantization switching signal exhibits non-linear quantization,
36. The image signal decoding apparatus according to claim 35, further comprising shift means for shifting the output from the multiplication means by a bit determined by the first quantization information. 40. The image signal decoding apparatus according to claim 39, wherein said multiplication means is composed of a three-stage adder.