Automatic evaluation system, method and equipment for video coding quality and storage medium

Abstract

The embodiment of the invention provides an automatic evaluation system, method and device for video coding quality and a storage medium, and the system comprises a cloud configuration center which is used for configuring a to-be-evaluated task name, a corresponding encoder type and live broadcast coding parameters required by the encoder of the type; the live broadcast stream signal generator is used for generating a labeled live broadcast stream; the encoder control module is used for receiving at least one group of live broadcast encoding parameters and the labeled live broadcast stream and generating at least one instruction; the live broadcast transcoding module is used for receiving the instruction, selecting a specified encoder and carrying out live broadcast transcoding according to live broadcast coding parameters; the live broadcast recording module is used for recording the transcoded live broadcast stream; and the full-reference evaluation engine is used for determining an evaluation index according to the recorded live broadcast stream and the initial video file, so that a live broadcast environment can be truly simulated, and the encoding effects of different types of live broadcast encoders in various scenes such as different video content types and different encoding parameters can be comprehensively and objectively evaluated.

Description

Technical Field

[0001] The present application relates to the technical field of video coding quality evaluation, and in particular to a system, method, device and storage medium for automatically evaluating video coding quality. Background Art

[0002] With the rapid development of live streaming and streaming services over the Internet, video encoders play a crucial role in ensuring video transmission quality and bandwidth utilization. Existing encoder evaluation methods are primarily based on on-demand encoders, which fail to fully reflect the performance of live encoders.

[0003] In the live streaming field, due to the continuity of live content and the need for perfect frame-level alignment of the preceding and following content, there is no established method for accurately evaluating the encoding quality of live encoders. Other existing evaluation methods primarily reference the evaluation of on-demand encoders. This involves transcoding on-demand files according to the parameters of the live encoder and then comparing the quality of the on-demand files before and after transcoding to evaluate the encoder's performance. While this method can achieve frame-level alignment of the preceding and following content to a certain extent, it also has the following drawbacks and cannot fully reflect the true performance of the live encoder.

[0004] First, it lacks coverage for real-world scenarios: Evaluating live encoders by encoding on-demand files cannot fully capture real-world live encoding scenarios. Live encoding is a continuous, lengthy process, requiring a fixed number of frames per second. The encoder must complete the encoding of each frame within strict time constraints. The encoding process for on-demand files, on the other hand, is more flexible. Depending on available computing resources, the encoder can mobilize powerful computing power to complete the transcoding of the entire file in a shorter timeframe. This difference prevents on-demand encoding evaluation methods from accurately simulating the actual operating environment of live encoders.

[0005] In live streaming scenarios, encoders need to process continuous data streams from other video sources in real time and must encode and output each frame within a fixed timeframe to ensure real-time and smooth video streaming. This real-time requirement places greater pressure on live encoders when processing data frames, especially when network bandwidth fluctuates or computing resources are limited. Encoders must balance bitrate control with image quality. On-demand encoders, on the other hand, can fully analyze the content of the entire on-demand file during the encoding process, optimizing the encoding of each frame and achieving higher image quality at the same bitrate. Therefore, evaluating live encoders based on on-demand files cannot truly reflect their performance in real-world applications.

[0006] Second, there are inconsistencies in the implementation of encoders for live and on-demand streaming: There are significant differences in the implementation of encoders for live and on-demand streaming, which also leads to limitations in evaluating live encoders using on-demand files. During live streaming, data frames are sent at a fixed frequency, and the encoder can only refer to limited previous frame information when processing each frame. This limitation makes live encoders face greater challenges in bitrate control and picture quality optimization. To ensure the real-time and smoothness of the video stream, live encoders usually need to make a trade-off between bitrate and image quality. Especially when computing resources are limited, the encoder may reduce image quality to maintain a stable bitrate.

[0007] During the encoding process of a VOD file, the encoder can reference the entire file content and comprehensively evaluate the encoding performance of each frame. This global optimization enables the VOD encoder to achieve better bitrate compression. Furthermore, VOD encoders can utilize more complex encoding algorithms, such as multi-pass encoding and scene change detection, to further improve encoding performance. Therefore, evaluating a live encoder using VOD files may yield better performance than a real-world encoder, but this does not truly reflect the performance of a live encoder in real-world applications.

[0008] Third, the limitations of evaluation methods: Existing evaluation methods primarily rely on transcoding comparisons of on-demand files. While these methods can achieve frame-level alignment, they cannot fully evaluate the performance of live encoders in real-world applications. Live encoders must handle a variety of complex scenarios, such as network bandwidth fluctuations, computing resource constraints, and real-time requirements. These factors significantly impact encoder performance. The encoding process for on-demand files cannot simulate these complex application scenarios.

[0009] In summary, there is an urgent need for an automated evaluation solution for video encoding quality that can realistically simulate a live broadcast environment, support multi-dimensional parameter configuration, and ensure frame-level alignment of the evaluation content. Summary of the Invention

[0010] The embodiments of the present application provide a system, method, device, and storage medium for automatically evaluating video encoding quality.

[0011] In a first aspect of an embodiment of the present application, a system for automatically evaluating video encoding quality is provided, comprising a cloud configuration center, a live stream signal generator, an encoder control module, a live transcoding module, a live recording module, and a full-reference evaluation engine, wherein the cloud configuration center, the live stream signal generator, and the live transcoding module are all connected to the encoder control module, and the live transcoding module and the full-reference evaluation engine are both connected to the live recording module.

[0012] The cloud configuration center is used to configure the name of the task to be evaluated, the corresponding encoder type, and the live encoding parameters required by the encoder type;

[0013] A live stream signal generator, configured to receive and segment the initial video file to be evaluated into labeled live streams;

[0014] An encoder control module is configured to receive at least one live broadcast encoding parameter and a labeled live broadcast stream, and generate at least one live broadcast transcoding instruction;

[0015] The live transcoding module is used to receive live transcoding instructions, select a specified encoder and perform live transcoding according to live encoding parameters;

[0016] The live streaming recording module is used to receive and record the transcoded live streaming;

[0017] A full-reference evaluation engine is used to receive the recorded live stream and determine evaluation metrics based on the recorded live stream and the original video file.

[0018] In an optional embodiment of the present application, in the cloud configuration center, the live broadcast encoding parameters include at least one parameter of the encoding standard, resolution, bit rate, GOP structure and number of B frames.

[0019] In an optional embodiment of the present application, in the live stream signal generator, a timing compensation amount is minimized by a preset compensation mechanism, wherein the timing compensation amount is obtained by the following expression:

[0020]

[0021] Among them, Δt is the timing compensation amount, is the time sequence of the live stream of the initial video file to be evaluated, It is a time series of labeled live stream.

[0022] In an optional embodiment of the present application, in the encoder control module, when multiple tagged live streams are received, the multiple tagged live streams are sorted according to the timestamps of the tags in the multiple tagged live streams, and for each sorted tagged live stream, a live transcoding task corresponding to the current tagged live stream is generated according to the current tagged live stream and its corresponding live encoder configuration item to be evaluated.

[0023] In an optional embodiment of the present application, in the live broadcast recording module, a double-buffered ring queue structure is used to record the transcoded live broadcast stream.

[0024] In an optional embodiment of the present application, in the full-reference evaluation engine, multi-scale spatial feature decomposition is performed on the recorded live stream and the initial video file respectively, and a real-time comparison of the two videos is performed through a multi-threaded parallel computing architecture to generate an evaluation report, wherein the evaluation report includes quantifiable indicators of multiple dimensions and quality fluctuation curves of the time dimension corresponding to each quantifiable indicator.

[0025] In an optional embodiment of the present application, the quantifiable indicators include peak signal-to-noise ratio, structural similarity and video multi-method evaluation fusion index, wherein,

[0026] The peak signal-to-noise ratio is given by the following expression:

[0027]

[0028] Where I is the pixel of the frame of the original video file, K is the pixel of the frame of the live stream after recording, M×N is the frame size, MAX I is the maximum value of the pixel,

[0029] The structural similarity is obtained by the following expression:

[0030]

[0031] Among them, μ x , μ y are the mean values ​​of the pixels of the original video file and the live stream after recording, are the variance of the pixels of the original video file and the live stream after recording, σ xy is the covariance of the pixels of the original video file and the live stream frame after recording, SSIM is the structural similarity, C1 and C2 are preset values,

[0032] The video multi-method evaluation fusion index is obtained by the following expression:

[0033] VMAF=F(f;θ)

[0034] Among them, f is the feature vector of the frame of the initial video file and the live stream after recording, θ is the model parameter, VMAF is the video multi-method evaluation fusion index, and F is the trained deep learning model based on multi-feature fusion.

[0035] A second aspect of the embodiments of the present application provides a method for automatically evaluating video coding quality based on an automated evaluation system for video coding quality, comprising:

[0036] Utilize the cloud configuration center to receive and provide the configuration items of the live encoder to be evaluated according to the parameter name of the live encoder to be evaluated;

[0037] Using a live stream signal generator to receive and split the initial video file to be evaluated into labeled live streams;

[0038] Using an encoder control module to receive at least one live broadcast encoding parameter and a labeled live broadcast stream, and generate at least one live broadcast transcoding instruction;

[0039] Use the live transcoding module to receive live transcoding instructions, select a specified encoder and perform live transcoding according to the live encoding parameters;

[0040] Use the live broadcast recording module to receive and record the transcoded live stream;

[0041] The full-reference evaluation engine receives the recorded live stream and determines the evaluation metrics based on the recorded live stream and the original video file.

[0042] According to a third aspect of an embodiment of the present application, a computer device is provided, comprising: a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the above-mentioned method for automatically evaluating video encoding quality are implemented.

[0043] According to a fourth aspect of the embodiments of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned method for automatically evaluating the quality of video encoding are implemented.

[0044] The above technical solutions provided by the embodiments of the present application have at least some or all of the following advantages compared to the prior art:

[0045] The automated evaluation system for video encoding quality described in the embodiment of the present application includes a cloud configuration center, a live stream signal generator, an encoder control module, a live transcoding module, a live recording module and a full-reference evaluation engine. The cloud configuration center is used to configure the name of the task to be evaluated and the corresponding encoder type and the live encoding parameters required by the encoder of this type; the live stream signal generator is used to receive and divide the initial video file to be evaluated into labeled live streams; the encoder control module is used to receive at least one live encoding parameter and labeled live stream, generate at least one live transcoding instruction, and generate at least one live transcoding task; the live transcoding module is used to receive the live transcoding instruction, select a specified encoder and perform live transcoding according to the live encoding parameters; the live recording module is used to receive and record the transcoded live stream; the full-reference evaluation engine is used to receive the recorded live stream and determine the evaluation indicators based on the recorded live stream and the initial video file, so as to comprehensively and objectively evaluate the encoding effect of the live encoder in various scenarios such as different video content types and different encoding parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0047] Figure 1 A structural diagram of an automated video coding quality evaluation system provided in one embodiment of the present application;

[0048] Figure 2 A flowchart of a method for automatically evaluating video encoding quality provided by one embodiment of the present application;

[0049] Figure 3 A schematic diagram of the computer device structure provided for one embodiment of the present application. DETAILED DESCRIPTION

[0050] In order to make the technical solutions and advantages of the embodiments of the present application more clearly understood, the exemplary embodiments of the present application are further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, and are not an exhaustive list of all the embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other unless they conflict.

[0051] See Figure 1 The automated video encoding quality evaluation system provided in the embodiment of the present application includes a cloud configuration center, a live stream signal generator, an encoder control module, a live transcoding module, a live recording module, and a full reference evaluation engine. The cloud configuration center, the live stream signal generator, and the live transcoding module are all connected to the encoder control module, and the live transcoding module and the full reference evaluation engine are all connected to the live recording module.

[0052] The cloud configuration center is used to configure the name of the task to be evaluated, the corresponding encoder type, and the live encoding parameters required by the encoder type;

[0053] A live stream signal generator, configured to receive and segment the initial video file to be evaluated into labeled live streams;

[0054] An encoder control module is configured to receive at least one live encoding parameter and a labeled live stream, generate at least one live transcoding instruction, and generate at least one live transcoding task;

[0055] The live transcoding module is used to receive live transcoding instructions, select a specified encoder and perform live transcoding according to live encoding parameters;

[0056] The live streaming recording module is used to receive and record the transcoded live streaming;

[0057] A full-reference evaluation engine is used to receive the recorded live stream and determine evaluation metrics based on the recorded live stream and the original video file.

[0058] In an optional embodiment of the present application, data interaction is achieved between the cloud configuration center, the live stream signal generator, the encoder control module, the live transcoding module, the live recording module and the full reference evaluation engine through the HTTP API (a standardized interface for transmitting data between the client and the server through the HTTP / HTTPS protocol).

[0059] In an optional embodiment of the present application, in the cloud configuration center, the live encoding parameters include at least one parameter of the encoding standard, resolution, bit rate, GOP structure (starting from the I frame (key frame), including several P frames (forward prediction frames) and B frames (bidirectional prediction frames), forming a cyclic structure such as IBBPBPBBPBBI) and the number of B frames. The cloud configuration center serves as a unified parameter management platform for the system, supporting centralized storage and dynamic updates of multiple encoder configurations. The configuration pulling component of the cloud configuration center can flexibly select and pull the corresponding encoding parameter set according to different test requirements, and realize flexible combination of multi-dimensional parameters such as different encoding standards (H264 / H265), resolutions (1080p / 720p / 480p), bit rates, GOP structures (IPPP / IBBP), and the number of B frames. The parameter parser of the cloud configuration center automatically converts the pulled parameters into a standardized XML structure, which is convenient for automatic recognition and calling of subsequent modules, thereby improving the scalability and compatibility of the system.

[0060] This application introduces a cloud-configurable multi-dimensional parameter matrix. Compared with the traditional evaluation method of transcoding on-demand files using fixed parameters, it significantly improves the flexibility of parameter management and the maintainability of the system. Through cloud configuration, users can dynamically adjust more than 20 encoding parameters according to actual needs, such as bit rate, resolution, frame rate, encoding algorithm, etc., so as to more accurately simulate the performance of encoders in different live broadcast scenarios. This flexible parameter management method can not only cover more complex application scenarios, but also quickly optimize the evaluation plan according to real-time needs, greatly improving the comprehensiveness and accuracy of the evaluation. In addition, cloud configuration also supports historical records and version management of parameters, which facilitates the backtracking and analysis of evaluation results under different parameter combinations, further enhancing the maintainability and scalability of the system, and can provide a more efficient and flexible solution for the performance evaluation of live encoders.

[0061] In actual applications, the live encoder of this application needs to process real-time input video streams, and comprehensively consider factors such as bit rate control, compression effect, and fluctuations in machine computing power. By simulating the input method of the live source stream, encoding is performed according to the specified data frame frequency to ensure that the evaluation process can truly reflect the performance of the live encoder.

[0062] This application uses multi-dimensional coverage of evaluation parameters and streaming encoding to enable the encoder to process the input video stream in a streaming manner and encode it frame by frame at a fixed frame rate, ensuring that the evaluation process is consistent with the actual live broadcast scenario and can comprehensively evaluate the encoder's performance in real-time processing capabilities, bit rate control and image quality optimization.

[0063] This application uses real-time bitrate control. During the evaluation process, the encoder dynamically adjusts the bitrate based on fluctuations in network bandwidth and machine computing power to ensure the smoothness of the video stream and the stability of image quality. This can comprehensively test the encoder's performance in dynamic bitrate control.

[0064] In an optional embodiment of the present application, the live stream signal generator not only supports the input of multiple video formats, but also realizes efficient format conversion and streaming media processing through the FFmpeg tool chain (FFmpeg tool chain is a complete set of cross-platform audio and video processing components, covering command line tools, development libraries and codec support, forming a multi-level workflow from basic operations to in-depth development). Through the FLV parser (FLV parser is a tool or program specifically used to parse FLV (FlashVideo, a streaming video container format developed by Adobe) format files, its core function is to disassemble the FLV file structure, extract metadata and audio and video data, and provide support for playback, transcoding or diagnosis) (FLV parser is only one of the methods, and this application is not limited to input in flv format), the initial video file to be evaluated is divided into tagged live streams, and dynamically issued according to the timestamp and type of the tag, thereby realizing an accurate simulation of the dynamic and real-time nature of streaming media data in a real live broadcast scenario. In the live stream signal generator, the timing compensation amount is minimized by a preset compensation mechanism, wherein the timing compensation amount is obtained by the following expression:

[0065]

[0066] Among them, Δt is the timing compensation amount, is the time sequence of the live stream of the initial video file to be evaluated, A time series of labeled live streams.

[0067] In this application, the live streaming signal generator minimizes the timing compensation amount through a preset compensation mechanism, and can simulate a real live streaming scenario.

[0068] To address the problem that traditional evaluation methods, which compare and transcode on-demand files, cannot simulate the real-time and continuity of live encoders in actual applications, this application uses real live source streams as input to simulate the entire live broadcast process, ensuring that the evaluation environment is completely consistent with the real scene:

[0069] This application uses dynamic tagged streaming media generation technology (FLV parser + real-time tag distribution) to use real live video content to simulate the live video source as the encoder input, rather than static on-demand files. This real live source stream input method can fully simulate the workflow of the live encoder in actual applications.

[0070] This application uses a real-time encoding process to enable the encoder to process the input video stream in real time at a fixed frame rate, ensuring that the evaluation process is consistent with the actual live broadcast scenario. In this way, the encoder's performance in real-time, continuity and stability can be comprehensively evaluated.

[0071] In an optional embodiment of the present application, the encoder control module, upon receiving multiple tagged live streams, sorts the streams according to the timestamps of the tags within the streams. For each sorted tagged live stream, a live transcoding task corresponding to the current tagged live stream is generated based on the current tagged live stream and its corresponding live encoder configuration item to be evaluated. The encoder control module interacts with upstream callers to implement various controls over the live transcoding module and can simultaneously manage the transcoding of multiple live streams.

[0072] In an optional embodiment of the present application, in the live broadcast recording module, a double-buffered ring queue structure is used to record the transcoded live broadcast stream to solve the timing alignment problem of the input and output of the live broadcast recording module.

[0073] In an optional embodiment of the present application, in the full-reference evaluation engine, multi-scale spatial feature decomposition is performed on the recorded live stream and the initial video file respectively, and a real-time comparison of the two videos is performed through a multi-threaded parallel computing architecture to generate an evaluation report, wherein the evaluation report includes quantifiable indicators of multiple dimensions and quality fluctuation curves of the time dimension corresponding to each quantifiable indicator.

[0074] In an optional embodiment of the present application, the quantifiable indicators include peak signal-to-noise ratio, structural similarity and video multi-method evaluation fusion index, wherein,

[0075] The peak signal-to-noise ratio is given by the following expression:

[0076]

[0077] Where I is the pixel of the frame of the original video file, K is the pixel of the frame of the live stream after recording, M×N is the frame size, MAX I is the maximum value of the pixel,

[0078] The structural similarity is obtained by the following expression:

[0079]

[0080] Among them, μ x , μ y are the mean values ​​of the pixels of the original video file and the live stream after recording, are the variance of the pixels of the original video file and the live stream after recording, σ xy is the covariance of the pixels of the original video file and the live stream frame after recording, SSIM is the structural similarity, C1 and C2 are preset values,

[0081] The video multi-method evaluation fusion index is obtained by the following expression:

[0082] VMAF=F(f;θ)

[0083] Among them, f is the feature vector of the frame of the initial video file and the live stream after recording, θ is the model parameter, VMAF is the video multi-method evaluation fusion index, and F is the trained deep learning model based on multi-feature fusion.

[0084] The system architecture of this application is highly modular, which facilitates functional expansion and maintenance upgrades. Through automated process design, it significantly reduces manual intervention and subjective errors, and improves the objectivity and consistency of the evaluation. The system can be widely used in live broadcast platforms, video cloud services, encoder R&D testing and other scenarios, providing the industry with standardized and automated encoder effect evaluation solutions.

[0085] Each module in the above system can be implemented in whole or in part through software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0086] See also Figure 2 The present application provides an automated evaluation method for video coding quality based on an automated evaluation system for video coding quality, comprising:

[0087] S210: Using the cloud configuration center, configure the task name to be evaluated, the corresponding encoder type, and the live encoding parameters required by the encoder type;

[0088] S220, using a live stream signal generator to receive and segment the initial video file to be evaluated into labeled live streams;

[0089] S230, using an encoder control module to receive at least one live broadcast encoding parameter and a tagged live broadcast stream, and generate at least one live broadcast transcoding instruction;

[0090] S240, using the live transcoding module to receive a live transcoding instruction, select a specified encoder, and perform live transcoding according to the live encoding parameters;

[0091] S250, using a live broadcast recording module to receive and record the transcoded live broadcast stream;

[0092] S260: Receive the recorded live stream using a full-reference evaluation engine, and determine evaluation indicators based on the recorded live stream and the initial video file.

[0093] The automated evaluation method for video encoding quality of this application can be widely used in video live broadcast platforms, streaming media service providers, video encoder development and testing and other fields, providing a scientific and reliable basis for encoder optimization and selection.

[0094] The automated evaluation method for video encoding quality in this application significantly reduces the need for manual intervention. Compared with the previous manual operation and process fragmentation of on-demand file evaluation, it can achieve full process automation, greatly improve evaluation efficiency and repeatability, and adapt to large-scale encoder version regression and multi-scenario evaluation needs.

[0095] For the specific definition of the above method, please refer to the definition of the automatic evaluation system for video encoding quality above, which will not be repeated here. It should be understood that although the various steps in the flowchart are displayed in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction for the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the figure may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0096] In one embodiment, a computer device is provided. The internal structure diagram of the computer device can be as follows: Figure 3As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, it realizes an automatic evaluation method for video encoding quality as described above. It includes: a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it realizes any step in the automatic evaluation method for video encoding quality as described above.

[0097] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, any step of the above-mentioned method for automatically evaluating video encoding quality can be implemented.

[0098] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0099] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0100] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0101] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0102] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0103] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. An automated video coding quality evaluation system, characterized in that: It includes a cloud configuration center, a live stream signal generator, an encoder control module, a live transcoding module, a live recording module and a full reference evaluation engine. The cloud configuration center, the live stream signal generator and the live transcoding module are all connected to the encoder control module, and the live transcoding module and the full reference evaluation engine are all connected to the live recording module. The cloud configuration center is used to configure the name of the task to be evaluated, the corresponding encoder type, and the live encoding parameters required by the encoder type; A live stream signal generator, configured to receive and segment the initial video file to be evaluated into labeled live streams; An encoder control module is configured to receive at least one live broadcast encoding parameter and a labeled live broadcast stream, and generate at least one live broadcast transcoding instruction; The live transcoding module is used to receive live transcoding instructions, select a specified encoder and perform live transcoding according to live encoding parameters; The live streaming recording module is used to receive and record the transcoded live streaming; A full-reference evaluation engine is used to receive the recorded live stream and determine evaluation metrics based on the recorded live stream and the original video file.

2. The system according to claim 1, wherein: In the cloud configuration center, the live broadcast encoding parameters include at least one parameter of encoding standard, resolution, bit rate, GOP structure and number of B frames.

3. The system according to claim 1, wherein: In the live stream signal generator, the timing compensation amount is minimized by a preset compensation mechanism, wherein the timing compensation amount is obtained by the following expression: Among them, Δt is the timing compensation amount, is the time sequence of the live stream of the initial video file to be evaluated, A time series of labeled live streams.

4. The system according to claim 1, wherein: In the encoder control module, when multiple tagged live streams are received, the multiple tagged live streams are sorted according to the timestamps of the tags in the multiple tagged live streams. For each sorted tagged live stream, a live transcoding task corresponding to the current tagged live stream is generated based on the current tagged live stream and its corresponding live encoder configuration item to be evaluated.

5. The system according to claim 1, wherein: In the live broadcast recording module, a double-buffered ring queue structure is used to record the transcoded live broadcast stream.

6. The system according to claim 1, wherein: In the full-reference evaluation engine, multi-scale spatial feature decomposition is performed on the recorded live stream and the initial video file respectively, and a real-time comparison of the two videos is performed through a multi-threaded parallel computing architecture to generate an evaluation report, wherein the evaluation report includes quantifiable indicators in multiple dimensions and quality fluctuation curves in the time dimension corresponding to each quantifiable indicator.

7. The system according to claim 1, wherein: The quantifiable indicators include peak signal-to-noise ratio, structural similarity and video multi-method evaluation fusion index, wherein, The peak signal-to-noise ratio is given by the following expression: Where I is the pixel of the frame of the original video file, K is the pixel of the frame of the live stream after recording, M×N is the frame size, MAX I is the maximum value of the pixel, The structural similarity is obtained by the following expression: Among them, μ x ,μ y are the mean values ​​of the pixels of the original video file and the live stream after recording, are the variance of the pixels of the original video file and the live stream after recording, σ xy is the covariance of the pixels of the original video file and the live stream frame after recording, SSIM is the structural similarity, C1 and C2 are preset values, The video multi-method evaluation fusion index is obtained by the following expression: VMAF=F(f;θ) Among them, f is the feature vector of the frame of the initial video file and the live stream after recording, θ is the model parameter, VMAF is the video multi-method evaluation fusion index, and F is the trained deep learning model based on multi-feature fusion.

8. A method for automatically evaluating video coding quality based on an automated evaluation system for video coding quality, characterized in that: include: Use the cloud configuration center to configure the task name to be evaluated, the corresponding encoder type, and the live encoding parameters required by the encoder type; Using a live stream signal generator to receive and split the initial video file to be evaluated into labeled live streams; Using an encoder control module to receive at least one live broadcast encoding parameter and a labeled live broadcast stream, and generate at least one live broadcast transcoding instruction; Use the live transcoding module to receive live transcoding instructions, select a specified encoder and perform live transcoding according to the live encoding parameters; Use the live broadcast recording module to receive and record the transcoded live stream; The full-reference evaluation engine receives the recorded live stream and determines the evaluation metrics based on the recorded live stream and the original video file.

9. A computer device comprising: The method comprises a memory and a processor, wherein the memory stores a computer program, and is characterized in that the processor implements the steps of the method according to claim 8 when executing the computer program.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to claim 8 are implemented.

Images