Low-power-consumption video decoding method and system for plug-in set top box

Abstract

The invention relates to the technical field of video decoding, and particularly provides a low-power-consumption video decoding method and system for a plug-in type set top box, and the method comprises the following steps: S1, after the plug-in type set top box receives a video stream, the video stream is preprocessed through a data preprocessing module; s2, integrating all links of real-time load monitoring and continuous tracking of a video decoding task in an operating system kernel of the plug-in set top box; s3, developing a highly adaptive drive program and an intelligent task allocation algorithm for a hardware acceleration unit commonly used in the plug-in set top box; and S4, designing a multi-layer cache architecture for the plug-in set top box. Compared with the prior art, the video decoding method has the advantages that the video data are preprocessed, the decoding efficiency is improved in a mode of combining software and hardware schemes, meanwhile, a dynamic voltage frequency adjustment and cache management optimization mechanism is added, and finally the purpose of reducing the video decoding power consumption is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of video decoding, and specifically provides a low-power video decoding method and system for a plug-in set-top box. Background Art

[0002] With its small size and convenient installation method, the plug-in set-top box has gradually emerged in the market. It is usually directly inserted into the HDMI interface or other compatible ports of the TV and powered by the TV without an additional power adapter. However, this power supply method determines that its available power is very limited.

[0003] When traditional video decoding technology is applied to a plug-in set-top box, the power consumption problem is more prominent: high power consumption not only easily causes the set-top box to overheat, affecting the stability of internal chips and components, but may also interfere with the normal operation of the TV and shorten the service life of the device. At the same time, due to the extremely small volume of the plug-in set-top box, the internal hardware resources are extremely limited, such as the computing power of the processor, the memory capacity, and the heat dissipation space are far less than those of a conventional set-top box. This makes it a key technical bottleneck that needs to be urgently broken through to achieve efficient and low-power video decoding under limited hardware conditions and power constraints when implementing the video decoding function. Summary of the Invention

[0004] The present invention aims at the above-mentioned deficiencies of the prior art and provides a practical low-power video decoding method for a plug-in set-top box.

[0005] A further technical task of the present invention is to provide a low-power video decoding system for a plug-in set-top box with reasonable design, safety and applicability.

[0006] The technical solution adopted by the present invention to solve its technical problems is as follows:

[0007] A low-power video decoding method for a plug-in set-top box has the following steps:

[0008] S1. After the plug-in set-top box receives the video stream, it first performs preprocessing through a data preprocessing module;

[0009] S2. Integrate real-time load monitoring into the operating system kernel of the plug-in set-top box to continuously track each link of the video decoding task;

[0010] S3. Develop highly adaptable driver programs and intelligent task allocation algorithms for the commonly used hardware acceleration units in the plug-in set-top box;

[0011] S4. Design a multi-layer cache architecture for the plug-in set-top box.

[0012] Further, in step S1, an algorithm that combines scene semantic analysis and pixel-level feature extraction is used to calculate the pixel differences and motion vectors between adjacent frames and intelligently identify the video scene type;

[0013] According to the hardware decoding capabilities of the plug-in set-top box and the characteristics of the video format, the input video data format is optimized and converted, and the complex video format is converted into a format that is more suitable for the hardware decoding of the set-top box during the preprocessing stage.

[0014] Further, in step S2, afterwards, with the help of hardware performance counters and a customized software monitoring program, the number of instruction executions and data processing rates of the processor under different decoding tasks are collected to build an accurate load model;

[0015] Based on the load model, when decoding simple video content such as low-resolution and low-frame-rate videos, the processor voltage and frequency are reduced to the lowest available level; when processing complex videos, they are dynamically increased to the appropriate level.

[0016] Further, in step S3, during the decoding process, according to the characteristics of different video coding formats and hardware acceleration units, the decoding tasks are intelligently allocated;

[0017] The algorithm for the software decoding part is deeply optimized. In motion estimation, an algorithm based on adaptive shrinkage of the search range is adopted, and the search range is dynamically adjusted according to the motion intensity of the video frame. At the same time, the memory access mode is optimized, and data prefetching and cache alignment technologies are used.

[0018] Further, in step S4, it includes on-chip high-speed cache, on-board memory cache, and cache using external storage expansion. According to the access frequency and timeliness of the video data, the frequently accessed key frame data and intermediate decoding results are stored in the on-chip high-speed cache; the relatively infrequently accessed but still quickly accessed data is stored in the on-board memory cache; the data for long-term storage and low access frequency is placed in the external storage cache;

[0019] A cache replacement strategy based on video content is adopted. According to the video playback order and scene switching, the data that may be accessed in the future is predicted. At the same time, combined with the data access frequency, the data that has not been accessed for a long time and has a lower importance is preferentially replaced.

[0020] A low-power video decoding system for a plug-in set-top box includes a data preprocessing module, a dynamic voltage and frequency adjustment module, a module combining hardware acceleration and software optimization, and a cache management and optimization module;

[0021] The data preprocessing module is used for the plug-in set-top box to receive the video stream and first perform preprocessing through the data preprocessing module;

[0022] The dynamic voltage and frequency adjustment module is used to integrate real-time load monitoring in the operating system kernel of the plug-in set-top box to continuously track all aspects of the video decoding task;

[0023] The hardware acceleration and software optimization combined module is used to develop highly adapted drivers and intelligent task allocation algorithms for commonly used hardware acceleration units in plug-in set-top boxes;

[0024] The cache management optimization module is used to design a multi-layer cache architecture for a plug-in set-top box.

[0025] Furthermore, the data preprocessing module uses an algorithm that integrates scene semantic analysis and pixel-level feature extraction to calculate pixel differences and motion vectors between adjacent frames and intelligently identify video scene types;

[0026] According to the hardware decoding capability and video format characteristics of the plug-in set-top box, the input video data format is optimized and converted, and the complex video format is converted into a format that is more suitable for the set-top box hardware decoding in the preprocessing stage.

[0027] Furthermore, the dynamic voltage and frequency adjustment module uses hardware performance counters and customized software monitoring programs to collect the number of instructions executed and the data processing rate of the processor under different decoding tasks to build an accurate load model;

[0028] Based on the load model, when decoding simple video content such as low resolution and low frame rate, the processor voltage and frequency are reduced to the lowest available level; when processing complex videos, they are dynamically increased to the adaptive level.

[0029] Furthermore, in the module combining hardware acceleration and software optimization, during the decoding process, decoding tasks are intelligently allocated according to the characteristics of different video encoding formats and hardware acceleration units;

[0030] The algorithm of the software decoding part is deeply optimized. In motion estimation, an algorithm based on adaptive contraction of the search range is adopted to dynamically adjust the search range according to the intensity of the video motion. At the same time, the memory access mode is optimized and data prefetching and cache alignment technology are used.

[0031] Furthermore, the cache management optimization module includes on-chip cache, on-board memory cache and cache expanded by external storage. According to the access frequency and timeliness of video data, frequently accessed key frame data and intermediate decoding results are stored in the on-chip cache; relatively infrequent but still required fast access data are stored in the on-board memory cache; long-term storage and low-frequency access data are placed in the external storage cache;

[0032] Adopt a cache replacement strategy based on video content. Predict the data that may be accessed in the future according to the video playback order and scene switching. At the same time, combined with the data access frequency, preferentially replace the data that has not been accessed for a long time and has a lower importance.

[0033] Compared with the prior art, an insertion type set-top box low-power video decoding method and system of the present invention has the following outstanding beneficial effects:

[0034] Through a series of innovative technical means, the present invention significantly reduces power consumption during the video decoding process of the insertion type set-top box, effectively alleviates the heating problem caused by power supply limitation, ensures the stable operation of the set-top box, and reduces the risk of equipment damage.

[0035] While reducing power consumption, utilize the synergistic effect of hardware acceleration and software optimization to ensure the smoothness and high quality of video decoding, providing an excellent viewing experience for users. Even under complex video content and limited hardware resources conditions, stable and efficient decoding can be achieved.

[0036] Low power consumption reduces the heating condition of the insertion type set-top box, reduces the risk of damage to hardware components caused by overheating, improves the stability and reliability of the equipment, extends the service life of the set-top box, and reduces the frequency of users replacing the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0038] Attached Figure 1 is a schematic flowchart of an insertion type set-top box low-power video decoding method;

[0039] Attached Figure 2 is a schematic flowchart of key frame extraction in an insertion type set-top box low-power video decoding method. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] In order to enable those skilled in the art to better understand the solution of the present invention, the following further detailed description of the present invention will be made in conjunction with specific embodiments. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0041] The following gives a preferred embodiment:

[0042] As Figure 1 , 2 shown, in this embodiment, a low-power video decoding method for a plug-in set-top box has the following steps:

[0043] S1. After the plug-in set-top box receives the video stream, it first performs preprocessing through a data preprocessing module;

[0044] Using an algorithm that combines scene semantic analysis and pixel-level feature extraction, accurately calculates the pixel differences and motion vectors between adjacent frames, and can intelligently identify video scene types, such as news interviews, sports events, movie blockbusters, etc. For relatively static scenes with little content change, such as news interview programs, further increase the redundant frame screening ratio and only retain key frames. The key frame selection algorithm comprehensively considers factors such as the amount of information in the picture, scene switching points, and changes in visual focus, ensuring that the selected key frames completely retain the core content of the video, greatly reducing the subsequent decoding data volume and effectively reducing power consumption.

[0045] According to the hardware decoding capabilities of the plug-in set-top box and the characteristics of the video format, optimize and convert the input video data format. In the preprocessing stage, convert complex video formats into formats that are more suitable for the fast decoding of the set-top box hardware, reducing the format parsing time and calculation amount during decoding, and further reducing power consumption.

[0046] S2. Integrate real-time load monitoring into the operating system kernel of the plug-in set-top box to continuously track all aspects of the video decoding task;

[0047] Including the computational loads in processes such as entropy decoding, motion compensation, and image reconstruction. Then, with the help of hardware performance counters and customized software monitoring programs, accurately collect key parameters such as the number of instruction executions and data processing rates of the processor under different decoding tasks, and build an accurate load model.

[0048] Based on the load model, formulate a refined multi-level dynamic voltage and frequency adjustment strategy. When decoding simple video content such as low-resolution and low-frame-rate videos, reduce the processor voltage and frequency to the lowest available level; when processing complex videos, such as high-definition and high-dynamic-range videos, dynamically increase to the appropriate level.

[0049] During the adjustment process, through the efficient interaction between the operating system and the hardware power management unit, achieve smooth switching of voltage and frequency, avoiding additional power consumption and system instability caused by frequent adjustments.

[0050] S3. Develop highly adaptable driver programs and intelligent task allocation algorithms for the commonly used hardware acceleration units in the plug-in set-top box;

[0051] During the decoding process, according to the characteristics of different video coding formats and hardware acceleration units, the decoding tasks are intelligently allocated. For example, tasks suitable for hardware parallel processing such as entropy decoding and inverse quantization are assigned to a dedicated decoding chip, while tasks such as complex post-processing of motion compensation are completed by an optimized software algorithm if the hardware acceleration effect is not good enough, giving full play to the advantages of hardware acceleration and reducing the load and power consumption of the processor.

[0052] Deeply optimize the algorithms in the software decoding part. In motion estimation, a fast algorithm based on adaptive shrinkage of the search range is adopted to dynamically adjust the search range according to the motion intensity of the video frame, reducing unnecessary calculations. At the same time, optimize the memory access mode, and use data prefetching and cache alignment technologies to reduce memory access latency, improve data processing efficiency, and further reduce power consumption during the software decoding process.

[0053] S4. Design a multi-layer cache architecture for the plug-in set-top box;

[0054] It includes on-chip high-speed caches (L1, L2 caches), on-board memory caches, and caches extended using external storage. According to the access frequency and timeliness of video data, the key frame data and intermediate decoding results that are frequently accessed are stored in the on-chip high-speed cache; data that is relatively less frequent but still needs to be accessed quickly is stored in the on-board memory cache; data that is stored for a long time and has a low access frequency is placed in the external storage cache. Through this hierarchical architecture, the cache hit rate is maximized, and the data reading time and power consumption are reduced.

[0055] Abandon the traditional general cache replacement algorithm and adopt a cache replacement strategy based on video content. Predict the data that may be accessed in the future according to the video playback order and scene switching. For example, when playing a serial drama, pre-cache the key data of the next episode; when playing a sports event, pre-cache video clips of exciting moments according to the progress of the event. At the same time, combined with the data access frequency, preferentially replace data that has not been accessed for a long time and has a lower importance to ensure that the cache always stores the most valuable data, improve the cache usage efficiency, and reduce power consumption.

[0056] Based on the above methods, a low-power video decoding system for a plug-in set-top box in this embodiment includes a data preprocessing module, a dynamic voltage and frequency adjustment module, a module combining hardware acceleration and software optimization, and a cache management optimization module;

[0057] The data preprocessing module is used to perform preprocessing on the video stream received by the plug-in set-top box through the data preprocessing module first;

[0058] Using an algorithm that combines scene semantic analysis and pixel-level feature extraction, calculate the pixel difference and motion vector between adjacent frames, and intelligently identify the video scene type;

[0059] According to the hardware decoding capabilities of the plug-in set-top box and the characteristics of video formats, optimize and convert the input video data format, and convert complex video formats into formats that are more suitable for the hardware decoding of the set-top box during the preprocessing stage.

[0060] The dynamic voltage and frequency adjustment module is used to integrate real-time load monitoring into the operating system kernel of the plug-in set-top box, and continuously track all aspects of the video decoding task;

[0061] With the help of hardware performance counters and customized software monitoring programs, collect the number of instruction executions and data processing rates of the processor under different decoding tasks, and build an accurate load model;

[0062] Based on the load model, when decoding simple video content such as low resolution and low frame rate, reduce the processor voltage and frequency to the lowest available level; when processing complex videos, dynamically increase it to the appropriate level.

[0063] The hardware acceleration and software optimization combination module is used to develop highly adaptable driver programs and intelligent task allocation algorithms for the commonly used hardware acceleration units in the plug-in set-top box;

[0064] During the decoding process, according to the characteristics of different video coding formats and hardware acceleration units, intelligently allocate decoding tasks;

[0065] Deeply optimize the algorithms for the software decoding part. In motion estimation, adopt an algorithm based on adaptive shrinkage of the search range, dynamically adjust the search range according to the motion intensity of the video picture, and at the same time, optimize the memory access mode, and use data prefetching and cache alignment technologies.

[0066] The cache management optimization module is used to design a multi-layer cache architecture for the plug-in set-top box;

[0067] It includes on-chip high-speed cache, on-board memory cache, and cache using external storage expansion. According to the access frequency and timeliness of video data, store frequently accessed key frame data and intermediate decoding results in the on-chip high-speed cache; relatively infrequently accessed but still requiring fast access data is stored in the on-board memory cache; data stored for a long time and with a low access frequency is placed in the external storage cache;

[0068] Adopt a cache replacement strategy based on video content, predict the data that may be accessed in the future according to the video playback order and scene switching, and at the same time, combine the data access frequency to preferentially replace data that has not been accessed for a long time and has a lower importance.

[0069] The above specific embodiments are only specific cases of the present invention. The patent protection scope of the present invention includes but is not limited to the above specific embodiments. Any technical solution that conforms to the technical solutions described in the above specific embodiments of the present invention and any appropriate changes or substitutions made by those of ordinary skill in the art shall fall within the patent protection scope of the present invention.

[0070] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A low-power video decoding method for a plug-in set-top box, characterized in that, The steps are as follows: S1. After the plug-in set-top box receives the video stream, it first performs preprocessing through the data preprocessing module; S2. Integrate real-time load monitoring into the operating system kernel of the plug-in set-top box to continuously track all aspects of the video decoding task; S3. Develop highly adaptable driver programs and intelligent task allocation algorithms for the commonly used hardware acceleration units in the plug-in set-top box; S4. Design a multi-layer cache architecture for the plug-in set-top box.

2. The method for low-power video decoding of a plug-in set-top box according to claim 1, characterized in that, In step S1, an algorithm that combines scene semantic analysis and pixel-level feature extraction is used to calculate the pixel differences and motion vectors between adjacent frames and intelligently identify the video scene type; According to the hardware decoding capabilities of the plug-in set-top box and the characteristics of the video format, the input video data format is optimized and converted, and the complex video format is converted into a format that better suits the hardware decoding of the set-top box during the preprocessing stage.

3. The plug-in set-top box low-power video decoding method according to claim 2, wherein, In step S2, then, with the help of hardware performance counters and customized software monitoring programs, the number of instruction executions and data processing rates of the processor under different decoding tasks are collected to build an accurate load model; Based on the load model, when decoding simple video content such as low resolution and low frame rate, the processor voltage and frequency are reduced to the lowest available level; when processing complex videos, it is dynamically increased to the appropriate level.

4. The plug-in set-top box low-power video decoding method according to claim 3, characterized in that In step S3, during the decoding process, according to the characteristics of different video coding formats and hardware acceleration units, the decoding tasks are intelligently allocated; The algorithms for the software decoding part are deeply optimized. In motion estimation, an algorithm based on adaptive shrinkage of the search range is adopted, and the search range is dynamically adjusted according to the motion intensity of the video picture. At the same time, the memory access mode is optimized, and data prefetching and cache alignment technologies are used.

5. The plug-in set-top box low-power video decoding method according to claim 4, wherein In step S4, it includes on-chip high-speed cache, on-board memory cache, and cache using external storage expansion. According to the access frequency and timeliness of video data, the frequently accessed key frame data and intermediate decoding results are stored in the on-chip high-speed cache; Relatively infrequently accessed but still requiring fast access data is stored in the on-board memory cache; data for long-term storage and low access frequency is placed in the external storage cache; A cache replacement strategy based on video content is adopted. According to the video playback order and scene switching, the data that may be accessed in the future is predicted. At the same time, combined with the data access frequency, the data that has not been accessed for a long time and has a lower importance is preferentially replaced.

6. An insertion type set-top box low-power video decoding system, characterized in that, It includes a data preprocessing module, a dynamic voltage and frequency adjustment module, a combination module of hardware acceleration and software optimization, and a cache management optimization module; The data preprocessing module is used for the plug-in set-top box to perform preprocessing through the data preprocessing module after receiving the video stream; The dynamic voltage and frequency adjustment module is used to integrate real-time load monitoring into the operating system kernel of the plug-in set-top box to continuously track all aspects of the video decoding task; The combination module of hardware acceleration and software optimization is used to develop highly adaptable driver programs and intelligent task allocation algorithms for the commonly used hardware acceleration units in the plug-in set-top box; The cache management optimization module is used to design a multi-layer cache architecture for the plug-in set-top box.

7. The plug-in set-top box low-power video decoding system according to claim 6, characterized in that, In the data preprocessing module, an algorithm that combines scene semantic analysis and pixel-level feature extraction is used to calculate the pixel difference and motion vector between adjacent frames and intelligently identify the video scene type; According to the hardware decoding capabilities of the plug-in set-top box and the characteristics of the video format, the input video data format is optimized and converted, and the complex video format is converted into a format that better suits the hardware decoding of the set-top box during the preprocessing stage.

8. An insertable set-top box low-power video decoding system according to claim 7, characterized in that, In the dynamic voltage and frequency adjustment module, with the help of hardware performance counters and customized software monitoring programs, the number of instruction executions and data processing rates of the processor under different decoding tasks are collected to build an accurate load model; Based on the load model, when decoding simple video content such as low-resolution and low-frame-rate videos, the voltage and frequency of the processor are reduced to the lowest available level; when processing complex videos, they are dynamically increased to the appropriate level.

9. The plug-in set-top box low-power video decoding system according to claim 8, characterized in that, In the module that combines hardware acceleration and software optimization, during the decoding process, decoding tasks are intelligently allocated according to the characteristics of different video coding formats and hardware acceleration units; The algorithms for the software decoding part are deeply optimized. In motion estimation, an algorithm based on adaptive shrinkage of the search range is adopted, and the search range is dynamically adjusted according to the motion intensity of the video frame. At the same time, the memory access mode is optimized, and data prefetching and cache alignment technologies are used.

10. An insertable set-top box low-power video decoding system according to claim 8, characterized in that, The cache management optimization module includes on-chip high-speed cache, on-board memory cache, and cache extended by external storage. According to the access frequency and timeliness of video data, the frequently accessed key frame data and intermediate decoding results are stored in the on-chip high-speed cache; Relatively infrequently accessed but still requiring fast access data is stored in the on-board memory cache; data for long-term storage and with low access frequency is placed in the external storage cache; A cache replacement strategy based on video content is adopted. According to the video playback order and scene switching, data that may be accessed in the future is predicted. At the same time, combined with the data access frequency, data that has not been accessed for a long time and has low importance is preferentially replaced.

Images