WO2020024160A1

WO2020024160A1 - Video processing device and method

Info

Publication number: WO2020024160A1
Application number: PCT/CN2018/098072
Authority: WO
Inventors: 陈秋伯; 郑萧桢
Original assignee: 深圳市大疆创新科技有限公司
Priority date: 2018-08-01
Filing date: 2018-08-01
Publication date: 2020-02-06
Also published as: CN110945870A

Abstract

Provided is a video processing method. The method comprises: calculating bandwidth consumption incurred in a video codec and related to reading and writing of external data; and determining, on the basis of the bandwidth consumption, a parameter associated with compression of a reference frame, the reference frame being applied to inter-frame prediction. The solution provided by the present application can be used to effectively meet storage bandwidth limitation of video codecs.

Description

Video processing device and method

Technical field

Embodiments of the present application relate to the field of video processing. More specifically, embodiments of the present application relate to a video processing device and a corresponding method.

Background technique

In order to reduce the bandwidth occupied by video storage and transmission, video data usually needs to be encoded and compressed. The encoding compression process includes prediction, transformation, quantization, and entropy encoding processes. Among them, prediction includes two types of intra prediction and inter prediction, the purpose of which is to remove redundant information of the current image block to be encoded by using prediction block information. Intra prediction uses the information of the frame image to obtain prediction block data, and inter prediction uses the information of the reference frame to obtain prediction block data.

Generally, a large amount of reference frame data is required for inter prediction, and these reference frame data are usually stored in an external memory. Reading and writing these reference frame data will consume a lot of bandwidth. With the continuous improvement of video resolution, the sharply increasing bandwidth requirements and the increase in memory power consumption have become urgent problems.

Summary of the invention

This application proposes that during the video encoding and decoding process, reference frame compression can be applied to compress the reference frame data, and the compression method and corresponding quantization parameters of the reference frame data are dynamically adjusted based on the bandwidth consumption of the video encoder / decoder to meet Bandwidth requirements and improve the quality of video images.

A first aspect of the embodiments of the present application provides a video processing method, including:

Calculate the bandwidth consumption associated with reading and writing external data in the video codec; and

Parameters associated with compression of a reference frame are determined based on the bandwidth consumption, the reference frame being used for inter prediction.

According to a second aspect of the embodiments of the present application, a video encoder / decoder is provided. The video codec includes a memory and one or more processors communicatively coupled with the memory. Instructions are stored on the memory, and when the instructions are executed by the one or more processors, cause the video codec to:

Calculating the bandwidth consumption related to reading and writing of external data in the video codec; and

According to a third aspect of the embodiments of the present application, there is provided a computer program that, when the computer program is run by at least one processor, causes the at least one processor to execute the method according to the first aspect of the embodiment of the present application. method.

According to a fourth aspect of the embodiments of the present application, there is provided a computer-readable storage medium storing the computer program according to the third aspect of the embodiments of the present application.

By adopting the embodiments of the present application, the memory bandwidth constraints of the video codec can be effectively satisfied, and the quality of the video image is improved.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and other features of the embodiments of the present application will become more apparent through the following detailed description in conjunction with the accompanying drawings, in which:

FIG. 1 is a schematic diagram illustrating a video codec scheme according to an embodiment of the present application.

FIG. 2 is a flowchart illustrating a video encoding and decoding method according to an embodiment of the present application.

FIG. 3 is a flowchart illustrating a method of determining a quantization parameter according to an embodiment of the present application.

FIG. 4 is a block diagram illustrating a video encoder / decoder according to an embodiment of the present application.

FIG. 5 is a schematic diagram illustrating a computer-readable storage medium according to an embodiment of the present application.

It should be noted that the drawings are not necessarily drawn to scale, and the focus is on illustrating the technical principles of the embodiments of the present application. In addition, for the sake of clarity, like reference numerals refer to like elements throughout the drawings.

detailed description

The technical solutions in the embodiments of the present application will be clearly described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the specification of the present application is only for the purpose of describing specific embodiments, and is not intended to limit the present application. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

As mentioned above, video codecs usually require a large amount of reference frame data for inter prediction. In video codec circuit design, these reference video frames are usually stored in external memory, so reading and writing these video reference frames will consume a lot of memory bandwidth. In order to meet the characteristics of video inter-frame prediction that randomly accesses a large number of reference frames, reference frame compression usually has the following characteristics:

a. It belongs to intra-frame compression, and only uses intra-frame information for residual prediction. Generally, the compression unit is very small, such as 4x4 or 8x8 pixels;

b. Read and write external memory with the image tile as the size. The image tile is generally composed of multiple compression units and the start address is determined or can be located by other information;

c. To support high parallel decoding throughput, it is usually necessary to store additional header information indicating the number of compressed bits per compression unit. The data stored in the external memory after the reference frame compression generally includes two parts: one is the compressed reference frame data, and the other is the reference frame header information. However, the header information is not limited to the number of compressed bits of the compression unit.

Reference frame compression can be integrated in the video encoder / decoder to effectively reduce the consumption of a large amount of bandwidth resources caused by reading and writing reference frame data. Reference frame compression can be divided into lossy compression and lossless compression. The advantage of lossless compression is to ensure that the reference frame data is exactly the same before encoding and after decoding, and the video encoding quality is not affected. The disadvantage of lossless compression is that the compression efficiency is low, and its compression rate changes dynamically with the video content, which causes the reference frame to read and write the external memory's bandwidth requirements also change dynamically and cannot be predicted. Generally speaking, bandwidth allocation needs to be considered in the worst case. The worst case of lossless reference frame compression is the corresponding bandwidth requirement when it is not compressed, so the bandwidth requirement cannot actually be less.

The advantage of lossy compression is that the compression efficiency is high. The compression rate and the bandwidth of accessing external memory can be guaranteed within a certain range through the control of quantization parameters. The disadvantage is that if the loss is too large, it will cause video encoding and decoding ends. Mismatch, affecting video image quality. Generally, lossy compression is performed in a fixed compression rate (for example, fixed to half of the original data amount). However, for reference frame compression units that are difficult to compress, the quality loss may be large due to the need to compress to a fixed size. In addition, the fixed compression rate does not consider the actual bandwidth consumption and cannot maximize the video encoding quality.

In view of the above, a method for adjusting the reference frame compression rate is proposed below. This method combines the characteristics of lossy and lossless compression, and dynamically adjusts whether to use lossy or lossless compression based on the bandwidth consumption of the video codec. , And the quantization parameters used in the case of lossy compression, so as to meet the storage bandwidth requirements of the video codec and improve the quality of the video image.

FIG. 1 is an overall schematic diagram showing a video codec scheme according to an embodiment of the present application. As shown in FIG. 1, the video encoder / decoder 10 includes a processing module 110, an entropy encoding module 120, a reference frame compression module 130, a reference frame decompression module 140, and a bandwidth consumption analysis module 160. Alternatively, the video codec 10 may further include a selectable input frame decompression module 170 (shown by a dashed box). The entropy encoding module 120, the reference frame compression module 130, the reference frame decompression module 140, and the optional input frame decompression module 170 are all capable of performing data read / write operations with the external memory 150.

The bandwidth consumption analysis module 160 is used to analyze the bandwidth consumption of the video encoder / decoder, and adjust the quantization parameter QP of the current image slice in the reference frame compression based on the analysis result, so as to dynamically adjust the reference frame compression in units of the image slice. Compression ratio.

Specifically, the bandwidth consumption analysis module 160 can monitor the bandwidth consumption modules in the video encoder / decoder. These modules have data interaction (read or write) with the external memory, which is the bandwidth consumption of the video encoder / decoder. The main source. As shown in Figure 1, there are four modules:

a) Reference frame compression module 130: used to compress the reconstructed image encoded by the video, wherein the image data is stored in the external memory 150 in a compressed form;

b) Input frame decompression module 170: The input video of the video encoder / decoder may be stored in the external memory 150 after frame compression, so the compressed video data read from the external memory 150 needs to be decompressed by the input frame decompression module and then sent to the video The codec performs video encoding. It should be noted that if this module does not exist, the input frame is read from the external memory 150 as the original data;

c) Reference frame decompression module 140: read the data compressed by the reference frame compression module 130 from the external memory 150, and perform decompression for inter prediction;

d) Entropy encoding module 120: used to generate an encoded code stream of a video encoder / decoder.

The bandwidth consumption analysis module 160 reads corresponding information from one or more of the four modules to analyze the bandwidth consumption. For example, when considering all these 4 modules, the total bandwidth consumption BW _all can be expressed by:

BW _all = BW _a + BW _b + BW _c + BW _d (1)

In the above formula, BW _a , BW _b , BW _c , and BW _d respectively represent the bandwidth consumption generated by the reference frame compression module 130, the input frame decompression module 170, the reference frame decompression module 140, and the entropy encoding module 120.

Preferably, for the BW _all calculated at any time point, the above components are statistics of the bandwidth consumption in the time period T from the current time point. For example, the bandwidth consumption of each module corresponding to the first L (corresponding to the T time period) video coding unit may be counted. The main reason to consider the bandwidth consumption in the T time period, not just the current time point, is to reduce the quality loss of the difficult-to-compress image area in order to meet the target bandwidth. In other words, the T time period is equivalent to considering the bandwidth constraint in a larger range, and the average compression ratio of the image in the larger time range is better than the compression ratio of the instantaneous difficult compression region. Therefore, the target bandwidth constraint is weakened, so that the overall image quality is improved.

It is noted that, although use of BW _a BW _all calculated in Equation _{_{(1), BW b, BW c, BW}} d four components, but with fewer components to calculate the BW _all, or may be added The other components that take up bandwidth consumption are used to calculate BW _all . For example, when only the bandwidth consumption of the reference frame compression module 130 and the reference frame decompression module 140 need to be considered, BW _all = BW _a + BW _c . In another example, the bandwidth consumption of the reference frame compression module 130 and the reference frame decompression module 140 and the entropy encoding module 120 may be considered, that is, BW _all = BW _a + BW _c + BW _d . Alternatively, the bandwidth consumption of the reference frame compression module 130 and the reference frame decompression module 140 and the input frame decompression module 170 may also be considered, that is, BW _all = BW _a + BW _b + BW _c .

The calculation of BW _a can be obtained by calculating the header information of the reference frame compression. Specifically, during the reference frame compression process, additional header information is stored while the compressed data is being stored for reference frame decompression. This header usually contains the number of compressed bits of an image slice. From this, the bandwidth consumption of an image slice can be calculated. Assuming that the writing of n image bands is completed within the time period T, the calculation formula of BW _a is:

In the above formula, bits_i is the number of compressed bits of the i-th image slice, which is provided by the header information of the corresponding image slice. The function bandwidth () converts the number of compressed bits into bandwidth consumption, which is related to the storage controller used by the system and its configuration parameters for reading and writing external memory.

For the calculation of BW _b , if there is no input frame decompression module 170, the bandwidth consumption is the amount of raw data read in the previous T period; if there is an input frame decompression module 170, the bandwidth consumption is read in the T period The bandwidth consumed by the input frame to compress the data.

For the calculation of BW _c , the bandwidth consumption is the bandwidth consumed by the reference frame compression data of multiple image strips read in the T period, which can be calculated using the above equation (2).

For the calculation of BW _d , the entropy coding module 120 can count the number of coding bits generated in the T time period and convert it into corresponding bandwidth consumption.

The above calculation of BW _all is used to dynamically adjust the compression rate of the current image slice in the reference frame compression module 130 so as to meet the target bandwidth limit BW _{codec of} the allocated video _codec . The BW _codec is the target bandwidth in the T period, which can be a fixed broadband or a target bandwidth that is dynamically allocated to the video codec at the system level. The target bandwidth BW _tar of the to-be-compressed image _slice of the current reference frame is:

BW _tar = BW _codec -BW _all (3)

In step S210, a bandwidth consumption related to reading and writing of external data in the video codec is calculated. Here, the “bandwidth consumption related to reading and writing of external data” refers to a bandwidth consumed by a read and write operation with an external memory (such as the external memory 150 in FIG. 1) (BW _all above). For example, as described above, the bandwidth consumption related to reading and writing of external data may include bandwidth consumption for reference frame compression and reference frame decompression. Alternatively, the bandwidth consumption related to external data reading and writing may also include the bandwidth consumption for input frames and / or entropy coding.

In one embodiment, calculating the bandwidth consumption related to the reading and writing of external data in the video codec may include calculating the bandwidth consumption within a predetermined time period T before the current time point. This can reduce the difficulty in compressing the image region from suffering a large quality loss in order to satisfy the target bandwidth.

In one embodiment, the bandwidth consumption for reference frame compression in the video codec (eg, BW _a above) may be calculated based on the header information of the reference frame compression.

In one embodiment, the bandwidth consumption (for example, BW _b above) of the input frame in the video encoder / decoder may be calculated based on the data amount of the input frame read in within the predetermined time period T.

In one embodiment, the bandwidth consumption (such as BW _c above) for decompressing the reference frame in the video encoder / decoder may be calculated based on the data amount of the reference frame compression data read in within a predetermined time period T.

In one embodiment, the bandwidth consumption (such as BW _d above) used for entropy coding in the video encoder / decoder may be calculated based on the data amount of the encoded data generated within the predetermined time period T.

Returning to FIG. 2, in step S220, parameters associated with compression of a reference frame are determined based on the bandwidth consumption, and the reference frame is used for inter prediction. For example, determining a parameter associated with reference frame compression may include determining a quantization parameter associated with a current image unit in the reference frame compression. The image unit may include, for example, an image strip.

In one embodiment, the value of the quantization parameter may be a predefined discrete value. The compression ratio of the reference frame can become higher as the quantization parameter becomes larger.

In one embodiment, the quantization parameter is determined such that the actual bandwidth consumption of the current image unit is not greater than the target bandwidth and is closest to the target bandwidth. For example, the target bandwidth may be calculated in the following manner:

1) obtaining a total bandwidth within the predetermined time period allocated for a video codec;

2) calculating a bandwidth consumption in the video codec within the predetermined time period; and

3) Calculate the difference between the total bandwidth and the bandwidth consumption as the target bandwidth.

Here, the total bandwidth can be dynamically allocated. Bandwidth consumption may include bandwidth consumption in a video codec for reference frame compression and reference frame decompression, and bandwidth consumption for input frames and / or entropy coding.

The method for determining the quantization parameter is described in detail below with reference to FIG. 3. As shown in FIG. 3, the method starts at block 300.

Assume that BW _real (QP) represents the actual bandwidth consumption of the image slice to be compressed, and QP is its corresponding quantization parameter. The compression ratio adjustment of the image band to be compressed is achieved by adjusting its quantization parameter QP. When QP is increased, the compression ratio becomes higher; when QP is decreased, the compression ratio becomes lower. Generally, QP values are discrete and can be predefined. For example, it can be set to QP _n (n = 0, 1, 2, 3 ...), where the smaller the value of _n , the smaller the QP _n , and QP ₀ represents lossless compression.

At block 310, it is determined whether BW _real (QP ₀ ) <= BW _tar holds. If it is, it means that lossless compression can meet the target bandwidth. At this time, the method proceeds to block 330 to obtain a quantization parameter QP = QP ₀ , and the current image strip uses lossless compression. If the determination at block 310 is no, the method proceeds to block 320.

At block 320, lossy compression needs to be employed to meet the target bandwidth. The quantization parameter QP of the current image slice is as small as possible, so that the actual bandwidth consumption is less than or equal to the QP of the target bandwidth. That is, QP = QPn is found such that BW _real (QP _n-1 )> BW _tar and BW _real (QPn) <= BW _tar .

At block 340, calculate the actual bandwidth consumption of the currently compressed image slice and update the BW _tar for the QP value calculation for the next image _slice to be compressed.

Finally, the method ends at block 350.

FIG. 4 is a block diagram illustrating a video encoder / decoder according to an embodiment of the present application. The video codec can be applied to a variety of platforms, such as drones, drones, or robots. As shown in FIG. 4, the video codec 40 includes a memory 410 and a processor 420.

The memory 410 stores program instructions. For example, the memory 410 may be a random access memory (RAM) or a read-only memory (ROM), or any combination thereof. The memory 410 may also include a persistent storage device such as any one or a combination of magnetic memory, optical memory, solid state memory, or even remotely mounted memory.

The processor 420 may include any combination of one or more of a central processing unit (CPU), a multi-processor, a microcontroller, a digital signal processor (DSP), an application specific integrated circuit, and the like.

The processor 420 may call a program instruction stored in the memory 410. When the program instructions are executed, the processor 420 may perform the following operations: calculate a bandwidth consumption related to reading and writing of external data in the video codec; and determine a bandwidth consumption associated with reference frame compression based on the bandwidth consumption Parameter, the reference frame is used for inter prediction.

In one example, the processor 420 may call program instructions stored in the memory 410. When the program instructions are executed, the processor 420 may perform the following operations: calculate a bandwidth consumption in a predetermined period of time before a current point in time. The bandwidth consumption related to reading and writing of external data may include bandwidth consumption for compression and decompression of reference frames in the video codec. Alternatively, the bandwidth consumption related to external data reading and writing may also include the bandwidth consumption for input frames and / or entropy coding in the video encoder / decoder.

In one example, the processor 420 may call program instructions stored in the memory 410. When the program instructions are executed, the processor 420 may perform the following operations: determine a quantization parameter associated with the current picture unit in the reference frame compression. The image unit may include an image strip.

In one example, the value of the quantization parameter may be a predefined discrete value. The compression ratio of the reference frame can become higher as the quantization parameter becomes larger.

In one example, the bandwidth consumption for reference frame compression in a video codec can be calculated based on the header information of the reference frame compression.

In one example, the bandwidth consumption for reference frame decompression in the video encoder / decoder may be calculated based on the data amount of the reference frame compression data read in within a predetermined time period T.

In one example, the bandwidth consumption for the input frame in the video encoder / decoder may be calculated based on the data amount of the input frame read in within the predetermined time period T.

In one example, the bandwidth consumption for entropy encoding in a video encoder / decoder may be calculated based on the data amount of the encoded data generated within the predetermined time period T.

In one example, the processor 420 may call program instructions stored in the memory 410. When the program instructions are executed, the processor 420 may perform the following operations: determine the quantization parameter so that the actual bandwidth consumption of the current image unit is not greater than the target bandwidth and is closest to the target bandwidth.

In one example, the processor 420 may call program instructions stored in the memory 410. When the program instructions are executed, the processor 420 may perform the following operations: obtain the total bandwidth allocated to the video codec within the predetermined time period; and calculate the video codec within the predetermined time period Bandwidth consumption; and calculating a difference between the total bandwidth and the bandwidth consumption as the target bandwidth.

In one example, the total bandwidth may be dynamically allocated. The bandwidth consumption may include bandwidth consumption in the video codec for reference frame compression and reference frame decompression. In addition, the bandwidth consumption may also include the bandwidth consumption in the video codec for input frames and / or entropy coding.

In addition, the embodiments of the present application may be implemented by means of a computer program product. For example, the computer program product may be a computer-readable storage medium. A computer program is stored on a computer-readable storage medium. When the computer program is executed on a computing device, related operations can be performed to implement the foregoing technical solutions of the embodiments of the present application.

For example, FIG. 5 is a block diagram illustrating a computer-readable storage medium 50 according to an embodiment of the embodiment of the present application. As shown in FIG. 5, the computer-readable storage medium 50 includes a computer program 510. The computer program 510, when executed by at least one processor, causes the at least one processor to perform various steps of, for example, the method described above in connection with FIG.

The computer program 510 stored on the computer-readable storage medium 50 may be loaded into the memory 410 of the video codec 40 shown in FIG. 4, for example, so that the processor 420 of the video codec 40 performs a corresponding operation.

Those skilled in the art can understand that examples of the computer-readable storage medium 50 include, but are not limited to, a semiconductor storage medium, an optical storage medium, a magnetic storage medium, or any other form of computer-readable storage medium.

The method and the related equipment of the embodiments of the present application have been described above with reference to the preferred embodiments. Those skilled in the art can understand that the method shown above is only exemplary. The method of the embodiment of the present application is not limited to the steps and sequence shown above. For example, the above steps may be performed in different steps from those in the embodiments of the invention, or may be performed in parallel.

It should be understood that the foregoing embodiments of the embodiments of the present application may be implemented by software, hardware, or a combination of both software and hardware. Such a setting of an embodiment of the present application is typically provided as software, code, and / or other data structures set or encoded on a computer-readable medium such as an optical medium (e.g., a CD-ROM), a floppy disk or hard disk, or One or more ROM or RAM or other media of firmware or microcode on a PROM chip, or downloadable software images, shared databases, etc. in one or more modules. Software or firmware or such a configuration may be installed on a computing device, so that one or more processors in the computing device execute the technical solutions described in the embodiments of the present application.

In addition, each functional module or individual feature of the device used in each of the above embodiments may be implemented or performed by a circuit, which is typically one or more integrated circuits. Circuits designed to perform the functions described in this specification may include general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs) or general-purpose integrated circuits, field-programmable gate arrays (FPGAs), or other Programming logic devices, discrete gate or transistor logic, or discrete hardware components, or any combination of the above. A general-purpose processor may be a microprocessor, or the processor may be an existing processor, controller, microcontroller, or state machine. The above-mentioned general-purpose processor or each circuit may be configured by a digital circuit, or may be configured by a logic circuit. In addition, when advanced technologies capable of replacing current integrated circuits appear due to advances in semiconductor technology, the embodiments of the present application may also use integrated circuits obtained using the advanced technologies.

The program running on the device according to the embodiment of the present application may be a program that causes a computer to realize the functions of the embodiment of the present application by controlling a central processing unit (CPU). The program or information processed by the program may be temporarily stored in volatile memory (such as random access memory RAM), hard disk drive (HDD), non-volatile memory (such as flash memory), or other memory systems. A program for implementing the functions of the embodiments of the present application may be recorded on a computer-readable recording medium. Corresponding functions can be realized by causing a computer system to read programs recorded on the recording medium and execute the programs. The so-called "computer system" herein may be a computer system embedded in the device, and may include an operating system or hardware (such as a peripheral device).

As described above, the embodiments of the present application have been described in detail with reference to the drawings. However, the specific structure is not limited to the above embodiments, and the embodiments of the present application also include any design changes that do not deviate from the gist of the embodiments of the present application. In addition, various modifications can be made to the description of the embodiments of the present application within the scope of the claims, and the embodiments obtained by appropriately combining the technical means of the different embodiments are also included in the technical scope of the embodiments of the present application. In addition, components having the same effects described in the above embodiments may be replaced with each other.

Claims

A video processing method includes:

Calculate the bandwidth consumption associated with reading and writing external data in the video codec; and

Parameters associated with compression of a reference frame are determined based on the bandwidth consumption, the reference frame being used for inter prediction.
The method according to claim 1, wherein calculating a bandwidth consumption related to reading and writing of external data in the video codec comprises calculating a bandwidth consumption in a predetermined time period before a current time point.
The method according to claim 1, wherein the bandwidth consumption related to reading and writing of external data in the video codec comprises: a bandwidth for compressing reference frames and decompressing reference frames in the video codec Consume.
The method according to claim 1 or 3, wherein the bandwidth consumption related to reading and writing of external data in the video codec further comprises: inputting frames and / or entropy in the video codec Encoding bandwidth consumption.
The method of claim 1, wherein determining a parameter associated with reference frame compression comprises determining a quantization parameter associated with a current image unit in the reference frame compression.
The method according to claim 3, wherein a bandwidth consumption for reference frame compression in the video codec is calculated based on header information for reference frame compression.
The method according to claim 3, wherein a bandwidth consumption for decompressing a reference frame in the video encoder / decoder is calculated based on a data amount of the reference frame compression data read in during the predetermined period of time.
The method according to claim 4, wherein the bandwidth consumption for the input frame in the video codec is calculated based on the data amount of the input frame read in during the predetermined period of time.
The method according to claim 4, wherein a bandwidth consumption for entropy encoding in the video encoder / decoder is calculated based on a data amount of the encoded data generated within the predetermined period of time.
The method according to claim 5, wherein the image unit includes an image strip.
The method according to claim 5, wherein the value of the quantization parameter is a predefined discrete value.
The method according to claim 11, wherein the compression ratio of the reference frame becomes higher as the quantization parameter becomes larger.
The method according to claim 5, wherein the quantization parameter is determined such that the actual bandwidth consumption of the current image unit is not greater than the target bandwidth and is closest to the target bandwidth.
The method according to claim 13, wherein the target bandwidth is calculated according to the following manner:

Obtaining a total bandwidth within the predetermined time period allocated for a video codec;

Calculating bandwidth consumption in the video codec during the predetermined time period; and

Calculate a difference between the total bandwidth and the bandwidth consumption as the target bandwidth.
The method of claim 14, wherein the total bandwidth is dynamically allocated.
The method according to claim 14, wherein the bandwidth consumption comprises: bandwidth consumption in the video encoder / decoder for reference frame compression and reference frame decompression.
The method according to claim 14, wherein the bandwidth consumption comprises: bandwidth consumption for reference frame compression and reference frame decompression in the video codec, and input for use in the video codec Frame and / or entropy-encoded bandwidth consumption.
A video encoder / decoder includes:

Memory; and

One or more processors communicatively coupled to the memory,

Wherein, instructions are stored in the memory, and when the instructions are executed by the one or more processors, the video encoder / decoder is enabled:

Calculating the bandwidth consumption related to reading and writing of external data in the video codec; and

Parameters associated with compression of a reference frame are determined based on the bandwidth consumption, the reference frame being used for inter prediction.
The video encoder / decoder according to claim 18, wherein calculating bandwidth consumption related to reading and writing of external data in the video encoder / decoder comprises calculating bandwidth consumption in a predetermined time period before a current time point.
The video encoder / decoder according to claim 18, wherein the bandwidth consumption related to reading and writing of external data in the video encoder / decoder comprises: reference frame compression and reference in the video encoder / decoder Bandwidth consumption for frame decompression.
The video encoder / decoder according to claim 18 or 20, wherein the bandwidth consumption related to reading and writing of external data in the video encoder / decoder further comprises: an input frame in the video encoder / decoder And / or entropy encoding bandwidth consumption.
The video codec according to claim 18, wherein determining parameters associated with reference frame compression comprises determining quantization parameters associated with a current picture unit in the reference frame compression.
The video encoder / decoder according to claim 20, wherein the bandwidth consumption for the reference frame compression in the video encoder / decoder is calculated based on the header information of the reference frame compression.
The video encoder / decoder according to claim 20, wherein a bandwidth for decompressing a reference frame in the video encoder / decoder is calculated based on a data amount of reference frame compression data read in the predetermined period of time. Consume.
The video encoder / decoder according to claim 21, wherein the bandwidth consumption for the input frame in the video encoder / decoder is calculated based on the data amount of the input frame read in within the predetermined period of time.
The video encoder / decoder according to claim 21, wherein a bandwidth consumption for entropy encoding in the video encoder / decoder is calculated based on a data amount of the encoded data generated within the predetermined period of time.
The video codec according to claim 22, wherein the image unit includes an image slice.
The video codec according to claim 22, wherein the value of the quantization parameter is a predefined discrete value.
The video codec according to claim 28, wherein a compression rate of the reference frame becomes higher as the quantization parameter becomes larger.
The video codec according to claim 22, wherein the quantization parameter is determined such that the actual bandwidth consumption of the current image unit is not greater than the target bandwidth and is closest to the target bandwidth.
The video codec according to claim 30, wherein the target bandwidth is calculated in the following manner:

Obtaining a total bandwidth within the predetermined time period allocated for a video codec;

Calculating bandwidth consumption in the video codec during the predetermined time period; and

Calculate a difference between the total bandwidth and the bandwidth consumption as the target bandwidth.
The video codec according to claim 31, wherein the total bandwidth is dynamically allocated.
The video encoder / decoder according to claim 31, wherein the bandwidth consumption comprises: bandwidth consumption for reference frame compression and reference frame decompression in the video encoder / decoder.
The video encoder / decoder according to claim 31, wherein the bandwidth consumption comprises: bandwidth consumption for reference frame compression and reference frame decompression in the video encoder / decoder, and the video encoder / decoder Bandwidth consumption in input frames and / or entropy coding.
A computer program comprising instructions for performing a method according to any one of claims 1-17 when run on one or more processors.
A computer-readable storage medium storing a computer program according to claim 35.