CN104243988A - Video encoding and decoding method and device, method for transferring video bitstream and video bitstream - Google Patents

Abstract

The present application discloses a video encoding method, which includes: determining the level of the currently encoded image; and writing the level of the currently encoded image into a code stream sent to a video decoding end. The present application discloses a video decoding method, including: obtaining the level of the current decoded image from the code stream received from the video encoding end; determining whether to store the reconstructed image of the current decoded image into the image storage space according to the level of the currently decoded image, if To store the reconstructed image of the currently decoded image into the image storage space, according to the level of the currently decoded image, the level of the image stored in the image storage space, whether the image storage space is full, the difference between the currently decoded image and the image stored in the image storage space The distance determines whether images already stored in the image storage space are retained. The application also discloses a video coding device, a video decoding device, a method for transmitting a video code stream and a code stream. By applying the technical solution disclosed in this application, the limited image storage space can be effectively managed, a complex inter-image reference structure can be realized with less overhead, and the anti-error performance of video coding and decoding can be improved.

Description

Video coding and decoding method and device, method and video code stream transmission method

technical field

The present application relates to the technical field of video encoding and decoding, and in particular to a video encoding method and device, a video decoding method and device, a method for transmitting a video code stream, and a video code stream.

Background technique

Reference image management in video codec technology mainly involves two aspects:

1. Whether the currently decoded image needs to be stored for reference in subsequent images;

2. Do you need to continue storing the images that have already been stored?

In the existing H.264 and HEVC standards, the space for storing decoded images is defined as a decoded image buffer (DPB). The AVS standard defines the space for storing decoded images as an image buffer. The image buffer is used to refer to it uniformly below.

H.264 uses a first-in-first-out (FIFO) management mechanism or a memory management command (MMCO) to manage image buffers. HEVC uses a buffer description mechanism (buffer description) to manage image buffers.

The first-in-first-out management mechanism is characterized by simplicity and effectiveness, but when the reference relationship is too complicated, such as the reference structure of level B, due to the characteristic of first-in-first-out in order, the most relevant one or several reference images cannot be remain in the image buffer. Both the memory management instruction and the buffer description mechanism are very flexible management mechanisms. Based on these two mechanisms, the image buffer can flexibly select the image to be stored. However, since the storage of reference images still has certain rules, that is, the images with stronger correlation (for example, those with high reconstruction quality, or referenced by many subsequent images) are more valuable to be stored. Too flexible selection will inevitably cost some money. necessary price.

In addition, for the management method of memory management instructions, the relationship between the decoded image buffers of two adjacent images is very close. If the data of a certain image is lost, the decoded image buffer of this image cannot be managed correctly, then Will affect the management of the decoded image buffer for all subsequent images.

Contents of the invention

The purpose of this application is to provide a level-based video encoding method and device, video decoding method and device to effectively manage the limited image storage space and realize complex image processing with less overhead. A reference structure between images.

The present application also provides a method for transmitting a video code stream and a video code stream, which can transmit image level information in the video code stream, so that in the case of loss of certain image data, it is guaranteed that the image decoded after the image The decoded image buffer can be properly managed to improve error resistance performance.

A video encoding method provided by this application includes:

Determine the level of the currently encoded image;

Write the level of the currently encoded image into the code stream sent to the video decoder.

Preferably, the method for determining the level of the currently encoded image includes: determining the level of the currently encoded image according to the management result of the image storage space.

Preferably, after determining the level of the current coded picture, it further includes: determining whether to store the reconstructed picture of the current coded picture in the image storage space according to the level of the current coded picture, if the reconstructed picture of the current coded picture is stored in the picture storage space , then according to the level of the currently encoded image, the level of the image stored in the image storage space, whether the image storage space is full, and the distance between the currently encoded image and the image stored in the image storage space, determine whether the image stored in the image storage space is reserve.

Preferably, if the level of the current coded picture is the lowest level, the reconstructed picture of the current coded picture is not stored in the picture storage space.

Preferably, the manner of determining whether the image stored in the image storage space is retained includes:

From the images stored in the image storage space, find out the image whose distance to the current coded image is negative and whose level is smaller than the level of the current coded image, in the found image, the absolute value of the distance from the current coded image The largest image is removed from the image storage space;

Or, from the images stored in the image storage space, find out the images whose distance to the current coded picture is negative and whose level is less than or equal to the level of the current coded picture, and among the found pictures, the distance from the current coded picture The image with the largest absolute value is deleted from the image storage space;

Or, find the image whose distance with the current coded image is negative from the images stored in the image storage space, and among the found images, the image with the largest absolute value of the distance from the current coded image is removed from the image storage space delete.

Preferably, writing the level of the currently coded image into the code stream sent to the video decoding end includes: writing at least one reference image set in the sequence header code stream, and each reference image set includes at least one level; Level reference information is written in the picture header stream of the coded picture, and the level reference information specifies a level in a set of reference pictures, which is the level of the current coded picture.

A video decoding method provided by this application includes:

Obtain the level of the current decoded image from the code stream received from the video encoder;

Determine whether to store the reconstructed image of the current decoded image into the image storage space according to the level of the currently decoded image. level, whether the image storage space is full, and the distance between the currently decoded image and the images stored in the image storage space determine whether the stored images in the image storage space are retained.

Preferably, if the level of the currently decoded picture is the lowest level, the reconstructed picture of the current decoded picture is not stored in the picture storage space.

Preferably, the manner of determining whether the image stored in the image storage space is retained includes:

From the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is smaller than the level of the current decoded image, in the found image, the absolute value of the distance from the current decoded image The largest image is removed from the image storage space;

Or, from the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is less than or equal to the level of the current decoded image, and in the found image, the distance to the current decoded image The image with the largest absolute value is deleted from the image storage space;

Or, find out the image whose distance to the current decoded image is negative from the images stored in the image storage space, and among the found images, remove the image with the largest absolute value from the image storage space from the image storage space delete.

Preferably, the obtaining the level of the currently decoded image from the code stream received from the video encoding end includes: obtaining at least one reference image set from the sequence header code stream, each reference image set includes at least one level; The level reference information is obtained from the picture header stream of the picture, and the level reference information specifies a level in a set of reference pictures, and the level is the level of the currently decoded picture.

A video encoding device provided by the present application includes at least an encoding processing unit and a code stream writing unit, wherein:

The encoding processing unit is used to determine the level of the currently encoded image;

The code stream writing unit is used for writing the level of the current coded image into the code stream sent to the video decoder.

Preferably, the encoding processing unit determines the level of the currently encoded image according to the management result of the image storage space.

Preferably, after the encoding processing unit determines the level of the current encoded image, it is further used to determine whether to store the reconstructed image of the current encoded image in the image storage space according to the level of the current encoded image, if the reconstructed image of the currently encoded image Stored in the image storage space, the image storage space is determined according to the level of the current encoded image, the level of the image stored in the image storage space, whether the image storage space is full, the distance between the current encoded image and the image stored in the image storage space Whether stored images are retained.

Preferably, if the level of the currently encoded image is the lowest level, the encoding processing unit does not store the reconstructed image of the currently encoded image in the image storage space.

Preferably, the encoding processing unit is configured to determine whether the image stored in the image storage space is retained in the following manner:

From the images stored in the image storage space, find out the image whose distance to the current coded image is negative and whose level is smaller than the level of the current coded image, in the found image, the absolute value of the distance from the current coded image The largest image is removed from the image storage space;

Or, from the images stored in the image storage space, find out the images whose distance to the current coded picture is negative and whose level is less than or equal to the level of the current coded picture, and among the found pictures, the distance from the current coded picture The image with the largest absolute value is deleted from the image storage space;

Or, find the image whose distance with the current coded image is negative from the images stored in the image storage space, and among the found images, the image with the largest absolute value of the distance from the current coded image is removed from the image storage space delete.

Preferably, the code stream writing unit is used to write at least one reference image set in the sequence header code stream, each reference image set includes at least one level; Writing level reference information, the level reference information designates a level in a set of reference pictures, and the level is the level of the current coded picture.

A video decoding device provided by the present application includes a decoding processing unit and a level information acquisition unit set in the code stream reading unit, wherein:

The level information obtaining unit is used to obtain the level of the currently decoded image from the code stream received from the video encoding end;

The decoding processing unit is used to determine whether to store the reconstructed image of the current decoded image into the image storage space according to the level of the currently decoded image, and if the reconstructed image of the currently decoded image is stored in the image storage space, then according to the level of the currently decoded image, The level of images stored in the image storage space, whether the image storage space is full, and the distance between the currently decoded image and the images stored in the image storage space determine whether the images stored in the image storage space are retained.

Preferably, if the level of the currently decoded picture is the lowest level, the decoding processing unit does not store the reconstructed picture of the current decoded picture in the picture storage space.

Preferably, the decoding processing unit is configured to determine whether the image stored in the image storage space is retained in the following manner:

From the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is smaller than the level of the current decoded image, in the found image, the absolute value of the distance from the current decoded image The largest image is removed from the image storage space;

Or, from the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is less than or equal to the level of the current decoded image, and in the found image, the distance to the current decoded image The image with the largest absolute value is deleted from the image storage space;

Or, find out the image whose distance to the current decoded image is negative from the images stored in the image storage space, and among the found images, remove the image with the largest absolute value from the image storage space from the image storage space delete.

Preferably, the level information acquiring unit is used to acquire at least one reference image set from the sequence header stream, each reference image set includes at least one level; and is used to acquire the level from the image header stream of the currently decoded image Reference information, the level reference information specifies a level in a set of reference pictures, and the level is the level of the currently decoded picture.

A method for transmitting a video code stream provided by this application includes:

Transmit at least one reference image set in the sequence header stream, each reference image set includes at least one level; transmit level reference information in the image header stream of the current image, and the level reference information specifies a reference image set A level of , which is the level of the current image.

Preferably, the order of levels included in each set of reference pictures is the decoding order of the corresponding pictures.

The present application also provides a video code stream, wherein the sequence header code stream contains at least one reference image set, and each reference image set includes at least one level; the image header stream of the current image contains level reference information, so The level reference information mentioned above specifies a level in a set of reference images, which is the level of the current image.

Preferably, the order of levels included in each set of reference pictures is the decoding order of the corresponding pictures.

It can be seen from the above technical solution that the technical solution provided by the present application can effectively manage a limited image storage space and realize a complex reference structure between images. At the same time, compared with the memory management instruction scheme and the buffer description mechanism, unnecessary overheads in transmission, storage, and processing brought about by an overly flexible reference image management scheme are reduced.

Moreover, the present application reserves appropriate required reference images for complex inter-image references, and has high code stream efficiency and certain anti-error performance.

Description of drawings

FIG. 1 is a schematic flow diagram of a preferred video encoding method of the present application;

Fig. 2 is a schematic flow chart of a preferred video decoding method of the present application;

FIG. 3 is a schematic diagram of the composition and structure of a preferred video encoding device of the present application;

FIG. 4 is a schematic diagram of the composition and structure of a preferred video decoding device of the present application.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and examples.

The main idea of the present application is to provide a technical solution for managing reference images based on levels.

In this application, when defining the level, the level of the current encoded/decoded image can be determined according to the correlation between the current encoded/decoded image and the subsequent encoded/decoded image or other standards, and the image is stored according to the level of the current encoded/decoded image Space is managed.

However, such a situation is not ruled out: assuming that a certain image storage space management result needs to be obtained at the encoding end, at this time, the level of the current encoded image can also be deduced from the management result of the image storage space, that is: according to the image storage space The result of the management determines the level of the currently encoded picture. In this way, after the level of the currently encoded image is obtained, since the image storage space management result has been determined, there is no need to manage the image storage space according to the level of the currently encoded image.

In this application, at the encoding end, the image storage space generally refers to all the buffers used to store the reconstructed image of the currently encoded image; at the decoding end, the image storage space generally refers to all the buffers used to store the reconstructed image of the currently decoded image.

FIG. 1 is a schematic flowchart of a preferred video encoding method of the present application. In the schematic diagram, the level of the currently encoded image is determined first, and then the image storage space is managed according to the level as an example for illustration. The method includes the following steps:

Step 101: Determine the level of the currently encoded image.

Step 102: Manage the image storage space according to the level of the currently encoded image.

Specifically, it can be determined whether to store the reconstructed image of the current encoded image into the image storage space according to the level of the currently encoded image, and if the reconstructed image of the currently encoded image is stored in the image storage space, further according to the level of the currently encoded image, the image The level of the image stored in the storage space, whether the image storage space is full, and the distance between the currently encoded image and the image stored in the image storage space determine whether the image stored in the image storage space is retained.

Here, the distance between the current encoded image and the images stored in the image storage space is determined according to the display order of the current encoded image and the images stored in the image storage space.

If the level of the currently coded picture is the lowest level, the reconstructed picture of the current coded picture is not stored in the picture storage space. Here, it is assumed that the levels are set sequentially. If the levels are set in reverse order, then when the level of the current encoded image is the highest level, the reconstructed image of the current encoded image will not be stored in the image storage space. Of course, in specific applications, the levels can also be set out of order, but it is necessary to determine which level or levels of images do not need to be stored in the image storage space, so that when the level of the currently encoded image falls within the range of these levels, it is not stored in the image storage space Stores the reconstructed image of the current encoded image in .

The above methods for determining whether the images stored in the image storage space are reserved include:

From the images stored in the image storage space, find out the image whose distance to the current coded image is negative and whose level is smaller than the level of the current coded image, in the found image, the absolute value of the distance from the current coded image The largest image is removed from the image storage space;

Or, from the images stored in the image storage space, find out the images whose distance to the current coded picture is negative and whose level is less than or equal to the level of the current coded picture, and among the found pictures, the distance from the current coded picture The image with the largest absolute value is deleted from the image storage space;

Or, find the image whose distance with the current coded image is negative from the images stored in the image storage space, and among the found images, the image with the largest absolute value of the distance from the current coded image is removed from the image storage space delete.

Step 103: Write the level of the currently coded image into the code stream sent to the video decoder.

Specifically, at least one reference image set may be written in the sequence header stream, and each reference image set includes at least one level; and level reference information may be written in the image header stream of the current coded image, said The level reference information specifies a level in a set of reference pictures, which is the level of the currently coded picture.

In practical applications, the level information corresponding to the level can also be set, and the level information can be written into the code stream and transmitted to the video decoding end.

Fig. 2 is a schematic flow chart of a preferred video decoding method of the present application, the method comprising the following steps:

Step 201: Obtain the level of the currently decoded image from the code stream received from the video encoder.

Step 202: Manage the image storage space according to the level of the currently decoded image.

Specifically, it can be determined whether to store the reconstructed image of the currently decoded image into the image storage space according to the level of the currently decoded image, and if the reconstructed image of the currently decoded image is stored in the image storage space, further according to the level of the currently decoded image, the image The level of the image stored in the storage space, whether the image storage space is full, and the distance between the currently decoded image and the image stored in the image storage space determine whether the image stored in the image storage space is retained.

The specific implementation of this step is similar to the specific description of step 102 in Figure 1, the only difference being that what step 102 describes is the management of the image storage space in the encoding process, and this step is the management of the image storage space in the decoding process, so It is sufficient to replace “encoding” in step 102 with “decoding”, which will not be repeated here.

Fig. 3 is a schematic diagram of the composition and structure of a preferred video encoding device of the present application, the video encoding device at least includes an encoding processing unit and a code stream writing unit, wherein:

The encoding processing unit is used to determine the level of the currently encoded image;

The code stream writing unit is used for writing the level of the current coded image into the code stream sent to the video decoder.

Preferably, the encoding processing unit can determine the level of the currently encoded image according to the management result of the image storage space.

After the encoding processing unit determines the level of the current encoded image, it can be further used to determine whether to store the reconstructed image of the current encoded image in the image storage space according to the level of the currently encoded image, if the reconstructed image of the currently encoded image is stored in the image storage space storage space, then determine the stored image storage space according to the level of the current encoded image, the level of the image stored in the image storage space, whether the image storage space is full, the distance between the current encoded image and the image stored in the image storage space Whether the image is preserved.

If the level of the currently encoded image is the lowest level, the encoding processing unit does not store the reconstructed image of the currently encoded image in the image storage space.

Corresponding to the method shown in FIG. 1, the encoding processing unit may determine whether the image stored in the image storage space is retained in the following manner:

From the images stored in the image storage space, find out the image whose distance to the current coded image is negative and whose level is smaller than the level of the current coded image, in the found image, the absolute value of the distance from the current coded image The largest image is removed from the image storage space;

Or, from the images stored in the image storage space, find out the images whose distance to the current coded picture is negative and whose level is less than or equal to the level of the current coded picture, and among the found pictures, the distance from the current coded picture The image with the largest absolute value is deleted from the image storage space;

Or, find the image whose distance with the current coded image is negative from the images stored in the image storage space, and among the found images, the image with the largest absolute value of the distance from the current coded image is removed from the image storage space delete.

Preferably, the code stream writing unit can be used to write at least one reference image set in the sequence header code stream, each reference image set includes at least one level; Level reference information is written in , where the level reference information specifies a level in a reference image set, and the level is the level of the currently coded image.

Fig. 4 is a schematic diagram of the composition and structure of a preferred video decoding device of the present application, the video decoding device at least includes a decoding processing unit and a level information acquisition unit arranged in the code stream reading unit, wherein:

The level information obtaining unit is used to obtain the level of the currently decoded image from the code stream received from the video encoding end;

The decoding processing unit is used to determine whether to store the reconstructed image of the current decoded image into the image storage space according to the level of the currently decoded image, and if the reconstructed image of the currently decoded image is stored in the image storage space, then according to the level of the currently decoded image, The level of images stored in the image storage space, whether the image storage space is full, and the distance between the currently decoded image and the images stored in the image storage space determine whether the images stored in the image storage space are retained.

Preferably, if the level of the currently decoded picture is the lowest level, the decoding processing unit does not store the reconstructed picture of the current decoded picture in the picture storage space.

Preferably, the decoding processing unit is configured to determine whether the image stored in the image storage space is retained in the following manner:

From the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is smaller than the level of the current decoded image, in the found image, the absolute value of the distance from the current decoded image The largest image is removed from the image storage space;

Or, from the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is less than or equal to the level of the current decoded image, and in the found image, the distance to the current decoded image The image with the largest absolute value is deleted from the image storage space;

Or, find out the image whose distance to the current decoded image is negative from the images stored in the image storage space, and among the found images, remove the image with the largest absolute value from the image storage space from the image storage space delete.

Preferably, the level information acquisition unit can be used to acquire at least one reference image set from the sequence header stream, each reference image set includes at least one level; and is used to acquire from the image header stream of the currently decoded image Level reference information, where the level reference information specifies a level in a set of reference pictures, and the level is the level of the currently decoded picture.

In the technical solution provided by the application, an important information needs to be transmitted between the encoding end and the decoding end, that is, the level. For this reason, the application provides a video code stream and a method for transmitting the video code stream.

Wherein, the sequence header stream of the video code stream contains at least one reference image set, and each reference image set includes at least one level; the image header stream of the current image contains level reference information, and the level reference information specifies A level in a set of reference images, which is the level of the current image.

The method for transmitting a video code stream transmits at least one reference image set in the sequence header code stream, and each reference image set includes at least one level; transmits level reference information in the image header code stream of the current image, and the level reference The message specifies a level in a set of reference pictures that is the level of the current picture.

Preferably, the order of levels included in each set of reference pictures is the decoding order of the corresponding pictures.

The technical solution of the present application will be described in further detail below with examples. Since the encoding process corresponds to the decoding process, and relates to the technical solution for managing the image storage space claimed in this application, it also corresponds to the encoding process and the decoding process. Therefore, in the following embodiments, only the encoding process is taken as the example to illustrate.

Step 1: Define the level.

As mentioned above, when defining the level, the level of the current encoded image may be determined according to the correlation between the current encoded image and subsequent encoded images or other criteria. For example, the degree of correlation between the current coded picture and the subsequent coded picture can be determined according to factors such as the degree of quantization of the current coded picture and the number of times it is referenced by the post-coded picture, and the stronger the correlation, the higher the level of the coded picture. level.

Table 1 to Table 7 illustrate how to set the level of the current encoded image under several typical image structures, and list in detail the display sequence number of the image in the image storage area before encoding the current image and the image storage area after encoding the current image The display serial number of the image in the image is analyzed for the frame number of the image that needs to be stored in the image storage area under various image structures, thus visually indicating the beneficial effect that the technical solution of the present application can obtain.

Table 1

Table 1 is an example under the IPPPP image structure. In this example, the image storage space can store up to 4 frames of images.

Table 2

Table 2 is an example under the IBBP image structure. In this example, the image storage space can store up to 2 frames of images.

table 3

Table 3 is an example under the Hierarchical4 image structure. In this example, the image storage space can store up to 4 frames of images.

Table 4

Table 4 is an example under the Hierarchical8 image structure. In this example, the image storage space can store up to 4 frames of images.

table 5

Table 5 is an example under the Hierarchical8 image structure. In this example, the image storage space can store up to 4 frames of images.

Table 6

Table 6 is an example under the Hierarchical8 image structure. In this example, the image storage space can store up to 4 frames of images.

Table 7

Table 7 is an example under the Hierarchical16 image structure. In this example, the image storage space can store up to 4 frames of images.

Step 2: Manage the image storage space according to the level of the currently encoded image.

Specifically, level a>=level b>=level c can be set, and according to the level of the current encoded image, the level of the image stored in the image storage space, whether the image storage space is full, whether the current encoded image and the image storage space are full Factors such as the distance of the stored images determine how to manage the image storage space. The specific operation is as follows:

a) If the level of the currently encoded picture is higher than or equal to level a:

1) If the number of buffered images in the image storage space is less than the maximum number of images that can be stored in the image storage space, that is, the image storage space is not full, then the reconstructed image of the current encoded image is moved into the image storage space; otherwise,

2) The image storage space is full, that is: the number of buffered images in the image storage space is equal to the maximum number of images that can be stored in the image storage space, then perform the following operations:

Judging according to the display order of images, if there are buffered images before the current encoded image stored in the image storage space, check the level of these buffered images:

If there is a buffer image whose level is lower than that of the current coded image among these buffered images, remove the buffered image with the largest display order difference from the current coded image from the image storage space, and then move the reconstructed image of the current coded image into image storage space; otherwise,

Remove the buffered image with the largest display order difference between the current encoded image and the current encoded image in the image storage space, and then move the reconstructed image of the current encoded/decoded image into the image storage space.

If there is no buffered image before the current encoded image in the image storage space, remove the buffered image with the largest display order difference between the image storage space and the current encoded image from the image storage space, and then move the reconstructed image of the current encoded image into Image storage space.

b) If the level of the currently encoded image is lower than level a and higher than or equal to level b:

1) If the number of buffered images in the image storage space is less than the maximum number of images that can be stored in the image storage space, then move the reconstructed image of the current coded image into the image storage space; otherwise,

2) The number of buffered images in the image storage space is equal to the maximum number of images that can be stored in the image storage space, then the following operations are performed:

Judging according to the display order of images, if there are buffered images before the current encoded image stored in the image storage space, check the level of these buffered images:

If there is a buffered image whose level is lower than or equal to the level of the current encoded image among these buffered images, remove the buffered image with the largest display order difference from the currently encoded image from the image storage space, and then reconstruct the current encoded image The image is moved into the image storage space; otherwise,

Remove the buffered image with the largest display sequence difference between the current encoded image and the current encoded image in the image storage space, and then move the reconstructed image of the current encoded image into the image storage space.

If there is no buffer image before the current coded image in the image storage space, remove the buffered image with the largest difference between the image buffer and the display order of the current coded image from the image memory space, and then move the reconstructed image of the current coded image into the image storage.

c) If the current decoded image level is less than level b and higher than or equal to level c,

1) If the number of buffered images in the image storage space is less than the maximum number of images that can be stored in the image storage space, then move the reconstructed image of the current coded image into the image storage space; otherwise,

2) The number of buffered images in the image storage space is equal to the maximum number of images that can be stored in the image storage space, then perform the following operations:

Judgment is made according to the display order of the images. If there is a buffered image before the current encoded image in the image storage space, the buffered image with the largest display order difference between the current encoded image and the current encoded image in the image storage space is removed from the image Remove from the storage space, and then move the reconstructed image of the current coded image into the image storage space;

If there is no buffer image before the current coded image in the image storage space, remove the buffered image in the image buffer with the largest display order difference from the current coded image from the image storage space, and then move the reconstructed image of the current coded image into the image memory space.

d) If the level of the currently encoded image is lower than level c, keep the image storage space unchanged.

Since the relationship between the threshold levels abc set for comparison is: level a>=level b>=level c, the number of allowable classifications is not necessarily all 4. For example, when level a=level b>level c, the number of one category is 0, that is, it is divided into three categories.

Step 3: Write the level of the currently encoded image into the code stream sent to the video decoder

As mentioned above, the present application proposes to write at least one reference image set in the sequence header stream, each reference image set includes at least one level; write level reference information in the image header stream of the current image, the The level reference information specifies a level in a set of reference images, which is the level of the current image.

Examples of syntax elements in the sequence header code stream are as follows:

in:

The value of successive_picture_set_number_minus1 is the number of reference image sets minus 1;

successive_picture_set_size indicates the number of images in the reference image set i (i=0.1...successive_picture_set_number_minus1);

Value_id[j] indicates the level of the jth (j=0.1….successive_picture_set_size-1) frame image in the reference image set i. The transmission order of Value_id[j] is the decoding order of the corresponding images.

Examples of syntax elements in the image header stream are as follows:

if(successive_picture_set_num_minus1>0)

{successive_picture_set_idx}

Value_idx

in:

successive_picture_set_idx indicates the index of the reference picture set used;

Value_idx indicates the level using the Value_idx-th picture in the reference picture set. Standard-compliant code streams should satisfy: if two adjacent encoded/decoded images belong to the same set of reference relations, the Value_idx of the current encoded image should be equal to the Value_idx+1 of the previous encoded image.

Through the above-mentioned technical solution of the present application, a limited image storage space can be effectively managed, and a complex reference structure between images can be realized. At the same time, compared with the memory management instruction scheme and the buffer description mechanism, unnecessary overheads in transmission, storage, and processing brought about by an overly flexible reference image management scheme are reduced.

Moreover, the present application reserves appropriate required reference images for complex inter-image references, and has high code stream efficiency and certain anti-error performance.

The above is only a preferred embodiment of the application, and is not intended to limit the application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the application should be included in the application. within the scope of protection.

Claims (24)

1. A video coding method, characterized in that, comprising: Determine the level of the currently encoded image; Write the level of the currently encoded image into the code stream sent to the video decoder. 2. The method according to claim 1, characterized in that: The manner of determining the level of the currently encoded image includes: determining the level of the currently encoded image according to the management result of the image storage space. 3. The method according to claim 1, characterized in that: After determining the level of the current coded picture, it further includes: determining whether to store the reconstructed picture of the current coded picture into the image storage space according to the level of the current coded picture, if the reconstructed picture of the current coded picture is stored in the picture storage space, then The level of the encoded image, the level of the image stored in the image storage space, whether the image storage space is full, the distance between the current encoded image and the image stored in the image storage space determine whether the image stored in the image storage space is retained. 4. The method according to claim 3, characterized in that: If the level of the currently coded picture is the lowest level, the reconstructed picture of the current coded picture is not stored in the picture storage space. 5. The method according to claim 3, characterized in that: The manner of determining whether the image stored in the image storage space is reserved includes: From the images stored in the image storage space, find out the image whose distance to the current coded image is negative and whose level is smaller than the level of the current coded image, in the found image, the absolute value of the distance from the current coded image The largest image is removed from the image storage space; Or, from the images stored in the image storage space, find out the images whose distance to the current coded picture is negative and whose level is less than or equal to the level of the current coded picture, and among the found pictures, the distance from the current coded picture The image with the largest absolute value is deleted from the image storage space; Or, find the image whose distance with the current coded image is negative from the images stored in the image storage space, and among the found images, the image with the largest absolute value of the distance from the current coded image is removed from the image storage space delete. 6. The method according to any one of claims 1 to 5, characterized in that: The writing of the level of the current coded picture into the code stream sent to the video decoding end includes: writing at least one reference picture set in the sequence header code stream, each reference picture set includes at least one level; Level reference information is written in the header stream, and the level reference information specifies a level in a set of reference pictures, and the level is the level of the currently coded picture. 7. A video decoding method, characterized in that, comprising: Obtain the level of the current decoded image from the code stream received from the video encoder; Determine whether to store the reconstructed image of the current decoded image into the image storage space according to the level of the currently decoded image. level, whether the image storage space is full, and the distance between the currently decoded image and the images stored in the image storage space determine whether the stored images in the image storage space are retained. 8. The method of claim 7, wherein: If the level of the currently decoded picture is the lowest level, the reconstructed picture of the current decoded picture is not stored in the picture storage space. 9. The method of claim 7, wherein: The manner of determining whether the image stored in the image storage space is reserved includes: From the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is smaller than the level of the current decoded image, in the found image, the absolute value of the distance from the current decoded image The largest image is removed from the image storage space; Or, from the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is less than or equal to the level of the current decoded image, and in the found image, the distance to the current decoded image The image with the largest absolute value is deleted from the image storage space; Or, find out the image whose distance to the current decoded image is negative from the images stored in the image storage space, and among the found images, remove the image with the largest absolute value from the image storage space from the image storage space delete. 10. The method according to any one of claims 7 to 9, characterized in that: The obtaining the level of the current decoded image from the code stream received from the video encoding end includes: obtaining at least one reference image set from the sequence header code stream, each reference image set includes at least one level; from the image header of the current decoded image The level reference information is obtained from the code stream, and the level reference information specifies a level in a reference image set, and the level is the level of the currently decoded image. 11. A video encoding device, characterized in that it includes at least an encoding processing unit and a code stream writing unit, wherein: The encoding processing unit is used to determine the level of the currently encoded image; The code stream writing unit is used for writing the level of the current coded image into the code stream sent to the video decoder. 12. The video encoding device according to claim 11, characterized in that: The encoding processing unit determines the level of the currently encoded image according to the management result of the image storage space. 13. The video encoding device according to claim 11, characterized in that: After the encoding processing unit determines the level of the current encoded image, it is further used to determine whether to store the reconstructed image of the current encoded image into the image storage space according to the level of the current encoded image, if the reconstructed image of the currently encoded image is stored in the image storage space space, then determine the image stored in the image storage space according to the level of the current encoded image, the level of the image stored in the image storage space, whether the image storage space is full, and the distance between the current encoded image and the image stored in the image storage space Whether to keep. 14. The video encoding device according to claim 13, characterized in that: If the level of the currently encoded image is the lowest level, the encoding processing unit does not store the reconstructed image of the currently encoded image in the image storage space. 15. The video encoding device according to claim 13, characterized in that: The encoding processing unit is used to determine whether the image stored in the image storage space is reserved in the following manner: From the images stored in the image storage space, find out the image whose distance to the current coded image is negative and whose level is smaller than the level of the current coded image, in the found image, the absolute value of the distance from the current coded image The largest image is removed from the image storage space; Or, from the images stored in the image storage space, find out the images whose distance to the current coded picture is negative and whose level is less than or equal to the level of the current coded picture, and among the found pictures, the distance from the current coded picture The image with the largest absolute value is deleted from the image storage space; Or, find the image whose distance with the current coded image is negative from the images stored in the image storage space, and among the found images, the image with the largest absolute value of the distance from the current coded image is removed from the image storage space delete. 16. The video encoding device according to any one of claims 11 to 15, characterized in that: The code stream writing unit is used to write at least one reference image set in the sequence header code stream, each reference image set includes at least one level; and is used to write the level reference in the image header code stream of the current coded image information, the level reference information specifies a level in a set of reference pictures, and the level is the level of the currently coded picture. 17. A video decoding device, characterized in that it includes a decoding processing unit and a level information acquisition unit arranged in the code stream reading unit, wherein: The level information obtaining unit is used to obtain the level of the currently decoded image from the code stream received from the video encoding end; The decoding processing unit is used to determine whether to store the reconstructed image of the current decoded image into the image storage space according to the level of the currently decoded image, and if the reconstructed image of the currently decoded image is stored in the image storage space, then according to the level of the currently decoded image, The level of the image stored in the image storage space, whether the image storage space is full, and the distance between the currently decoded image and the image stored in the image storage space determine whether the image stored in the image storage space is retained. 18. The video decoding device according to claim 17, characterized in that: If the level of the currently decoded picture is the lowest level, the decoding processing unit does not store the reconstructed picture of the current decoded picture in the picture storage space. 19. The video decoding device according to claim 17, characterized in that: The decoding processing unit is used to determine whether the image stored in the image storage space is reserved in the following manner: From the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is smaller than the level of the current decoded image. In the found image, the absolute value of the distance from the current decoded image The largest image is removed from the image storage space; Or, from the images stored in the image storage space, find out the image whose distance to the current decoded image is negative and whose level is less than or equal to the level of the current decoded image, and in the found image, the distance to the current decoded image The image with the largest absolute value is deleted from the image storage space; Or, find out the image whose distance to the current decoded image is negative from the images stored in the image storage space, and among the found images, remove the image with the largest absolute value from the image storage space from the image storage space delete. 20. The video decoding device according to any one of claims 17 to 19, characterized in that: The level information acquiring unit is used to acquire at least one reference image set from the sequence header stream, each reference image set includes at least one level; and is used to acquire level reference information from the image header stream of the currently decoded image, so The level reference information mentioned above specifies a level in a set of reference pictures, which is the level of the currently decoded picture. 21. A method for transmitting a video stream, characterized in that: Transmit at least one reference image set in the sequence header stream, each reference image set includes at least one level; transmit level reference information in the image header stream of the current image, and the level reference information specifies a reference image set A level of , which is the level of the current image. 22. The method of claim 21, wherein: The order of levels included in each set of reference pictures is the decoding order of the corresponding pictures. 23. A video stream, characterized in that: The sequence header stream contains at least one reference image set, and each reference image set includes at least one level; the image header stream of the current image includes level reference information, and the level reference information specifies one of a reference image set Level, which is the level of the current image. 24. The method of claim 23, wherein: The order of levels included in each set of reference pictures is the decoding order of the corresponding pictures.