Distributed audio and video adaptive production system for multiple terminals

Abstract

The present invention discloses a distributed audio and video adaptive production system for multiple terminals, which relates to the field of audio and video production technology and includes a distributed storage cluster, audio and video processing nodes, a cloud scheduling center, a perception module, a content analysis module, and a client. The storage cluster uses erasure codes and Ceph, and an enhanced data redundancy rate formula ensures data reliability. The perception module collects terminal information to form a device feature vector, and the content analysis module uses an algorithm to generate a content feature vector. The cloud scheduling center allocates tasks based on the comprehensive processing capability adaptation factor. The client interface is adaptive and multi-threaded to improve efficiency, and interacts with the cloud scheduling center to meet the needs of different terminal users. The present invention uses distributed storage to ensure data security and layered storage to reduce costs. Cloud scheduling optimizes task allocation to improve efficiency. Intelligent perception and content analysis realize personalized production. Adaptive coding adapts to multiple terminals. Real-time monitoring ensures system stability. Open interfaces expand the ecosystem and enhance the audio and video production experience.

Description

Technical Field

[0001] The present invention relates to the field of audio and video production systems, and in particular to a distributed audio and video adaptive production system for multiple terminals. Background Art

[0002] With the rapid development of information technology, audio and video content is playing an increasingly important role in people's daily lives and work. From online education and video conferencing to live entertainment and short video creation, audio and video applications are becoming increasingly widespread. At the same time, the types of terminal devices are also becoming more diverse, including smartphones, tablets, laptops, and smart TVs. These devices vary significantly in hardware performance, screen resolution, and network connectivity. This places higher demands on audio and video production systems, requiring them to adapt to the characteristics of different terminals and provide high-quality audio and video content.

[0003] Traditional audio and video production systems primarily operate in a standalone environment, relying on local hardware resources for audio and video processing and editing. This approach has numerous limitations. First, processing power is limited by the hardware configuration of the local device. This often leads to performance bottlenecks for high-definition and ultra-high-definition audio and video production, as well as for complex special effects, resulting in slow processing and low production efficiency. Second, traditional systems lack the ability to adapt to multiple terminals. The playback effects of the produced audio and video content vary significantly across different terminals, making it difficult to meet the viewing needs of users across various devices.

[0004] With the development of cloud computing and distributed technologies, a number of cloud-based audio and video production systems have emerged. While these systems have addressed processing power and storage capacity issues to a certain extent, they still have some shortcomings. For example, in terms of data storage, most existing systems use simple redundant backup strategies that lack comprehensive consideration of data importance and access frequency, resulting in wasted storage resources. In terms of task allocation, they fail to fully consider the actual processing capacity of processing nodes and the correlation between tasks, which can easily lead to load imbalances and affect overall system performance.

[0005] Furthermore, current audio and video production systems lack intelligent adaptive encoding capabilities. While some systems can perform simple bitrate adjustments based on the device's network bandwidth, these systems fail to fully consider the device's screen resolution, processing power, and the characteristics of the audio and video content, failing to provide users with a personalized playback experience. Furthermore, existing systems also have security vulnerabilities, making audio and video data vulnerable to attacks and leaks during transmission and storage, posing a security risk to users.

[0006] Furthermore, with the development of the audio and video industry, more and more third-party developers hope to integrate audio and video production functions into their own applications. However, existing systems lack open API interfaces and a comprehensive developer support system, which limits the system's application scenarios and the development of the ecosystem. Therefore, the development of a distributed, adaptive audio and video production system for multiple terminals is of great practical significance. Summary of the Invention

[0007] The present invention proposes a distributed audio and video adaptive production system for multiple terminals to solve the problems mentioned in the above-mentioned prior art.

[0008] In order to achieve the above purpose, the present invention adopts the following technical solutions: a distributed audio and video adaptive production system for multiple terminals, including a distributed storage cluster, audio and video processing nodes, a cloud scheduling center, a perception module, a content analysis module and clients adapted to different terminals; the distributed storage cluster uses erasure coding technology combined with a distributed file system to store audio and video materials, and uses an enhanced data redundancy rate formula Ensure data reliability, including is the number of redundant data blocks, is the number of parity check blocks, α is the erasure code weight coefficient, which is dynamically adjusted according to the importance of data and storage environment; the perception module collects the hardware information, network status and user usage habits of the terminal device in real time, and uses the device feature vector To describe the terminal characteristics; the content analysis module uses learning algorithms to perform semantic analysis, scene recognition and emotion classification on audio and video materials to obtain content feature vectors , audio and video processing nodes are distributed in different geographical locations and have different processing capabilities. The cloud dispatch center adapts the factors according to the comprehensive processing capabilities. Assign audio and video processing tasks, including is the actual processing capacity of the node, The flexibility for nodes to handle different types of tasks, and is the weighting coefficient.

[0009] Furthermore, the following modules are also included:

[0010] Real-time monitoring module: The module collects the storage utilization of distributed storage clusters in real time ,in The used storage space, is the total storage space and the task completion rate of the audio and video processing nodes ,in is the number of completed tasks, is the total number of tasks and the energy consumption E of the node; the cloud scheduling center uses the data and energy consumption-performance ratio formula Dynamically adjust task allocation and storage strategies.

[0011] Adaptive coding module: Calculates adaptive coding parameters based on the terminal device's network bandwidth B, screen resolution R, processing power P, and content feature vector CFV ,in is the weight coefficient.

[0012] Furthermore, the real-time monitoring module also monitors the network delay L between the client and the cloud dispatch center, using the network delay fluctuation coefficient Assess network stability; is the fluctuation range of network delay, is the average network delay); when the network delay fluctuation coefficient exceeds the threshold, the cloud scheduling center adopts an adaptive buffer strategy and adjusts the buffer capacity through the formula Adjust the buffer capacity, where is the cushioning coefficient.

[0013] Furthermore, the cloud scheduling center considers the load balancing factor of the processing nodes when adjusting task allocation. and the task relevance coefficient; where is the processing capacity used by the i-th processing node, is the total processing capacity of the i-th processing node, and n is the number of processing nodes); the cloud scheduling center maintains a task dependency graph to record the relationship between each task, and calculates the task correlation coefficient by analyzing the input and output data and processing flow of the task; for related tasks, priority is given to nodes that are adjacent or have low communication costs, and the task allocation optimization formula is used. Evaluate allocation options.

[0014] Furthermore, the adaptive coding module also considers the complexity C of the audio and video content, and adjusts the adaptive coding parameters based on the complexity to ,in is the complexity adjustment factor.

[0015] Furthermore, the distributed storage cluster supports hierarchical storage of data, based on the access popularity of the data. ,in is the number of data accesses, Calculate the data storage priority for the statistical time period and the timeliness of the data T By storing data on storage media with different performance, and θ are weighting coefficients.

[0016] Furthermore, the client provides audio and video editing functions, supporting users to edit audio and video and add subtitles; during the editing process, the system will process the audio and video according to the processing power of the client device. The operation response time estimation formula is calculated based on the operation complexity OC. Optimize operational processes in advance, including is the response time coefficient.

[0017] Furthermore, the system uses encryption algorithms to encrypt audio and video data during transmission and storage, and uses the encryption strength evaluation formula Evaluate cryptographic security, where is the encryption key length, is the strength coefficient; in the data storage process, the audio and video materials and production projects are encrypted and stored, and the encryption strategy is dynamically adjusted in combination with the security level SL of the terminal device. The encryption adjustment formula is Ensure the privacy and security of audio and video data.

[0018] Furthermore, the system provides an open API interface, allowing third-party developers to integrate; through the API call success rate formula Evaluate API performance, where is the number of successful API calls, =The total number of API calls; if the API call success rate is low, the system will analyze the reasons and predict that the server load is too high or the interface parameters are incorrect, and make timely optimization and adjustments; the cooperation value index is calculated based on the usage frequency UF and data interaction volume DI of third-party applications. Provide differentiated services and support for different third-party applications, including and is the weighting coefficient

[0019] Compared with the existing technology, the beneficial effects of the present invention are:

[0020] For data storage, the distributed storage cluster utilizes erasure coding technology combined with a distributed file system, and ensures data reliability through an enhanced data redundancy formula. Furthermore, storage priority is calculated based on data popularity and timeliness, enabling tiered storage. This prioritizes frequently accessed and time-sensitive data on high-speed media, effectively reducing access latency and saving storage costs.

[0021] In task allocation, the cloud scheduling center allocates tasks based on the comprehensive processing capacity adaptation factor, while taking into account the load balancing factor and task correlation coefficient. Through the task allocation optimization formula, a more reasonable task allocation is achieved to avoid unbalanced load on processing nodes and improve the overall system performance and processing efficiency.

[0022] The adaptive coding module is a highlight of this system. It comprehensively considers the network bandwidth, screen resolution, processing power of the terminal device, as well as the feature vectors and complexity of the audio and video content, calculates adaptive coding parameters, and adopts layered coding technology to provide personalized coding solutions for different terminals, ensuring that audio and video can achieve the best playback effect on various terminals and improving user experience.

[0023] The real-time monitoring module collects real-time data from various aspects of the system, such as storage utilization, task completion rate, energy consumption, and network latency, and evaluates and analyzes it using relevant formulas. The cloud scheduling center dynamically adjusts task allocation, storage strategies, and buffer capacity based on this data to ensure stable system operation, improve resource utilization efficiency, and reduce energy consumption.

[0024] The client offers a wealth of audio and video editing features. A response time estimation formula is used to optimize workflows and reduce user wait times. An intelligent recommendation algorithm recommends appropriate editing templates and materials based on user habits and content feature vectors, improving editing efficiency and personalization.

[0025] The system excels in security, employing multiple encryption algorithms to encrypt audio and video data during transmission and storage. It dynamically adjusts encryption strategies based on the terminal's security level and uses an encryption strength assessment formula to ensure data privacy and security. Furthermore, the system offers an open API interface, evaluates performance based on API call success rates, and provides differentiated services for third-party applications based on a partnership value index, expanding the system's application scenarios and ecosystem. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a schematic block diagram of the distributed audio and video adaptive production system for multiple terminals proposed by the present invention.

[0027] Figure 2 This is a schematic block diagram of the distributed audio and video adaptive production method for multiple terminals proposed by the present invention. DETAILED DESCRIPTION

[0028] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0029] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0030] In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the said features. In the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined. In addition, the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be a connection between the two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. The present invention will be further described in detail below with reference to the accompanying drawings.

[0031] Reference Figures 1 to 2 :A distributed audio and video adaptive production system for multiple terminals, including a distributed storage cluster, multiple audio and video processing nodes, a cloud scheduling center, an intelligent perception module, a content analysis module, and clients adapted to different terminals. The distributed storage cluster uses erasure coding technology combined with a distributed file system (such as Ceph) to store audio and video materials, and uses an enhanced data redundancy rate formula ( is the number of redundant data blocks, is the number of parity check blocks, α is the erasure code weight coefficient, which is dynamically adjusted according to the importance of the data and the storage environment) to ensure data reliability. The intelligent perception module collects the hardware information of the terminal device (such as the number of CPU cores, GPU memory size), network status (including bandwidth, packet loss rate) and user usage habits (such as commonly used editing functions, viewing preferences) in real time, and uses the device feature vector The content analysis module uses deep learning algorithms to perform semantic analysis, scene recognition, and sentiment classification on audio and video materials to obtain content feature vectors. , used for subsequent personalized production. Multiple audio and video processing nodes are distributed in different geographical locations, with different processing capabilities. The cloud dispatch center adapts the factors according to the comprehensive processing capabilities. ( is the actual processing capacity of the node, The flexibility for nodes to handle different types of tasks, and The client interacts with the cloud dispatch center via the network, supporting devices with different operating systems and hardware configurations. The client interface adapts its layout to the device's screen size and resolution, and employs multi-threading technology to improve data transmission and processing efficiency. The distributed storage cluster uses erasure coding technology combined with the Ceph distributed file system to store audio and video materials. Using an enhanced data redundancy ratio formula, the α value is dynamically adjusted based on data importance and the storage environment, ensuring reliable material storage. The intelligent perception module collects real-time terminal device hardware information, network status, and user usage habits, comprehensively describing terminal characteristics using device feature vectors to provide a basis for personalized services. The content analysis module uses deep learning algorithms to analyze audio and video materials and derive content feature vectors to facilitate personalized production. Multiple audio and video processing nodes are distributed across different locations and have varying capabilities. The cloud dispatch center rationally allocates tasks based on the comprehensive processing capability adaptation factor. The client adapts to different terminals, with an adaptive interface layout. Multi-threading technology improves data transmission and processing efficiency. This ensures the reliable storage of audio and video materials, enables personalized services tailored to different terminals and user habits, and improves production efficiency through rational task allocation. Its excellent adaptability and efficient data processing capabilities deliver a high-quality user experience.

[0032] The present invention also includes the following modules:

[0033] Real-time monitoring module: This module collects the storage utilization of the distributed storage cluster in real time ( The used storage space, is the total storage space), task completion rate of audio and video processing nodes ( is the number of completed tasks, is the total number of tasks) and the energy consumption E of the node. The real-time monitoring module is deployed with high-precision sensors and data acquisition software, which collect data for the storage cluster and processing nodes respectively. For the storage cluster, the sensor monitors the read and write status, temperature and other parameters of the hard disk in real time, and combines the used and total storage space information recorded by the storage management software to accurately calculate the storage utilization rate. For audio and video processing nodes, by monitoring the operating status of the CPU and GPU and the energy consumption data of the power module, the task completion rate and energy consumption are accurately obtained. The cloud scheduling center uses this data and the energy consumption-performance ratio formula to calculate the performance of the node. Dynamically adjust task allocation and storage policies. When a processing node's energy-to-performance ratio is too high, the cloud scheduling center will migrate some tasks to nodes with lower energy-to-performance ratios. Simultaneously, based on storage utilization, it will migrate less frequently used data to low-cost storage media to balance system resource usage and energy consumption. Furthermore, the real-time monitoring module visualizes collected data, allowing administrators to monitor system status in real time and identify potential issues promptly.

[0034] Adaptive coding module: This module calculates adaptive coding parameters based on the network bandwidth B, screen resolution R, processing power P and content feature vector CFV of the terminal device ( is the weight coefficient). The adaptive encoding module uses a multi-model fusion method to calculate parameters. First, historical data is trained through a machine learning model to obtain the initial values ​​of different weight coefficients. In actual operation, the weight coefficient is dynamically adjusted according to the terminal device information and content feature vectors collected in real time. The audio and video processing node encodes the audio and video according to this parameter to adapt to the playback requirements and content characteristics of different terminals. For terminals with low network bandwidth, the adaptive encoding module will reduce the encoding bit rate and the amount of data. At the same time, it selectively retains key frames and feature information according to the importance of the content; for terminals with high-resolution screens, the clarity and details of the encoding are improved to ensure the playback effect. In addition, the adaptive encoding module also supports the conversion of multiple encoding formats, such as H.264, H.265, etc., and selects the appropriate encoding format according to the compatibility and performance of the terminal device.

[0035] In the present invention, the real-time monitoring module also monitors the network delay L between the client and the cloud dispatch center, using the network delay fluctuation coefficient ( is the fluctuation range of network delay, The real-time monitoring module uses the ping-pong mechanism to periodically send test data packets to the client and calculate the network delay by recording the time difference between sending and receiving. At the same time, the network delay data over a period of time is statistically analyzed to obtain the average network delay and delay fluctuation range. When the network delay fluctuation coefficient exceeds the threshold, the cloud scheduling center adopts an adaptive buffering strategy and adjusts the buffer capacity through the formula ( The Cloud Dispatch Center dynamically adjusts the buffer size on both the client and server sides. When network latency fluctuates significantly, the buffer size is increased to cache more data in advance to address potential latency spikes. Once the network stabilizes, the buffer size is appropriately reduced to alleviate data storage pressure. Furthermore, the Cloud Dispatch Center adjusts data transmission rates and priorities based on network latency, prioritizing the transmission of critical data, such as key frames of audio and video.

[0036] In the present invention, the cloud scheduling center considers the load balancing factor of the processing node when adjusting the task allocation ( is the processing capacity used by the i-th processing node, is the total processing capacity of the i-th processing node, n is the number of processing nodes) and the task correlation coefficient. The cloud scheduling center maintains a task dependency graph to record the relationship between each task. By analyzing the input and output data and processing flow of the task, the task correlation coefficient is calculated. For related tasks, priority is given to nodes that are adjacent or have low communication costs. The task allocation optimization formula is used. Comprehensively evaluate the allocation plan. When a new task arrives, the cloud scheduling center first calculates the load balancing factor of each processing node and the correlation coefficient between the task and existing tasks on the node. Then, based on the task allocation optimization formula, the cloud scheduling center selects the optimal processing node for task allocation. The cloud scheduling center also considers the network topology and communication bandwidth between nodes to minimize data transmission costs between tasks. If the load on a node is too high, the cloud scheduling center will migrate some tasks to other nodes with lower loads to achieve load balancing across the entire system.

[0037] In the present invention, the adaptive coding module also considers the complexity C of the audio and video content, and adjusts the adaptive coding parameters based on the complexity to ( (where is the complexity adjustment factor). The adaptive encoding module calculates the complexity of audio and video content by analyzing multiple dimensions, including frame rate, resolution variation, color complexity, and audio spectrum. For highly complex audio and video content, encoding quality requirements are appropriately increased. Layered encoding technology is also employed to selectively transmit data of different quality layers based on terminal capabilities. During layered encoding, audio and video data is divided into a base layer and an enhancement layer. The base layer contains essential audio and video information and is suitable for playback on low-performance terminals. The enhancement layer provides higher clarity, more detail, and better audio quality, suitable for high-performance terminals. The adaptive encoding module dynamically determines which quality layer to transmit based on the terminal's processing power and network bandwidth. For example, for terminals with limited network bandwidth, only the base layer data is transmitted; for terminals with sufficient network bandwidth and high processing power, the base layer and some enhancement layer data are transmitted. Furthermore, the adaptive encoding module adjusts encoding strategies based on user viewing preferences and historical behavior to provide a personalized playback experience.

[0038] In the present invention, the distributed storage cluster supports hierarchical storage of data, based on the access popularity of the data. ( is the number of data accesses, Calculate the data storage priority based on the statistical time period and the timeliness of the data ( and θ are weighting coefficients), storing data on storage media with different performance. The distributed storage cluster adopts a multi-level storage architecture, including a cache layer (such as SSD), a large-capacity storage layer (such as HDD), and an archive storage layer (such as a tape library). The storage management software regularly collects statistical data access times and times, calculates access popularity and timeliness, and stores the data in the appropriate storage layer according to the data storage priority. Frequently accessed and time-sensitive data is stored on high-speed storage media to reduce access latency; data with low access frequency but that needs to be stored for a long time is stored in the archive storage layer to save costs. At the same time, the storage management software monitors data access in real time and automatically migrates data to the appropriate storage layer when the access popularity or timeliness of the data changes. For example, when the access popularity of a piece of data suddenly increases, it is migrated from the large-capacity storage layer to the cache layer; when the timeliness of the data decreases and the access frequency decreases, it is migrated from the cache layer to the archive storage layer.

[0039] In the present invention, the client provides audio and video editing functions, supporting users to edit audio and video, add subtitles, etc. During the editing process, the system can process the audio and video according to the processing power of the client device. The operation response time estimation formula is calculated based on the operation complexity OC. ( Response time coefficient), pre-optimizing operational processes. The client's editing interface utilizes a modular design, with each editing module operating independently. Multi-threading technology is used to process user operations in parallel, improving editing efficiency. When a user performs an operation, the system first estimates the response time based on the complexity of the operation and the processing power of the client device. If the estimated response time is long, the system pre-loads and caches data, optimizing the algorithm execution sequence and reducing user wait time. Simultaneously, the client utilizes an intelligent recommendation algorithm to recommend appropriate editing templates and materials based on user usage habits and content feature vectors. Based on a deep learning model, this intelligent recommendation algorithm analyzes the user's historical operation data and audio and video content characteristics to identify user interests and preferences. When a user opens the editing interface, the system recommends commonly used editing templates and related materials, such as subtitle styles and music materials, based on the user's usage habits. Furthermore, the client supports user feedback on recommendation results, continuously optimizing the recommendation algorithm and improving the accuracy and personalization of recommendations.

[0040] In the present invention, the system uses an encryption algorithm to encrypt audio and video data during transmission and storage, and uses the encryption strength evaluation formula ( is the encryption key length, The system supports multiple encryption algorithms, such as AES and RSA, and selects the appropriate encryption algorithm based on the sensitivity of the data and the usage scenario. During data transmission, the SSL / TLS protocol is used to encrypt the data to ensure the security of the data during network transmission. During data storage, audio and video materials and production projects are encrypted and stored, and only authorized users can decrypt and access them. At the same time, the encryption strategy is dynamically adjusted based on the security level SL of the terminal device. The encryption adjustment formula is: , ensuring the privacy and security of audio and video data. For devices with higher security levels, a stronger encryption algorithm and longer encryption keys are used. For devices with lower security levels, encryption strength is appropriately reduced to improve system performance while ensuring a certain level of security. Furthermore, the system regularly updates encryption keys to prevent them from being cracked, further enhancing data security.

[0041] In this invention, the system provides an open API interface, allowing third-party developers to integrate. The success rate formula of API call is ( is the number of successful API calls, The system provides detailed API documentation and development examples to help third-party developers quickly integrate. The API interface adopts a RESTful architecture with good scalability and compatibility. The system monitors API calls in real time, records the request parameters, response time and results of each call, and evaluates the performance of the API based on the success rate of API calls. If the success rate of API calls is low, the system will analyze the reasons, which may be due to excessive server load, incorrect interface parameters, etc., and make timely optimization and adjustments. At the same time, the cooperation value index is calculated based on the usage frequency UF and data interaction volume DI of third-party applications. ( and is a weighting coefficient), providing differentiated services and support for different third-party applications, and expanding the application scenarios and ecosystem of the system. For third-party applications with a high cooperation value index, the system will provide higher service priority, more technical support and preferential policies; for third-party applications with a low cooperation value index, the system will guide them to optimize their usage and improve the value of cooperation. The system opens API interfaces for third-party developers to integrate, evaluates its performance through the API call success rate formula, and measures the interface operation status with intuitive data. At the same time, detailed API documentation and development examples are provided, just like preparing clear "instructions" and "operation guides" for developers to help them get started quickly with integration. The interface adopts a RESTful architecture, which is like building a stable and flexible framework with good scalability and compatibility, and can easily cope with diverse development needs.

[0042] The system monitors API calls in real time, recording request parameters and other information in detail. This acts as if a dedicated person is constantly monitoring operational dynamics. If a low call success rate is detected, the cause is quickly analyzed and optimized. Furthermore, a partnership value index is calculated based on the frequency of use and data interaction volume of third-party applications, enabling differentiated service provision. This initiative has numerous benefits, expanding the system's application scenarios and ecosystem while encouraging third-party application optimization, achieving a win-win situation for both parties and enhancing the overall competitiveness and practicality of the system.

[0043] The above are only preferred specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solutions and inventive concepts of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A distributed audio and video adaptive production system for multiple terminals, characterized by: It includes distributed storage cluster, audio and video processing nodes, cloud scheduling center, perception module, content analysis module and clients adapted to different terminals; the distributed storage cluster uses erasure coding technology combined with distributed file system to store audio and video materials, and uses enhanced data redundancy rate formula Ensure data reliability, including is the number of redundant data blocks, is the number of parity check blocks, α is the erasure code weight coefficient, and the α value is dynamically adjusted according to the importance of data and storage environment; the perception module collects the hardware information, network status and user usage habits of the terminal in real time, and uses the device feature vector To describe the terminal characteristics; The content analysis module uses learning algorithms to perform semantic analysis, scene recognition and emotion classification on audio and video materials to obtain content feature vectors. , audio and video processing nodes are distributed in different geographical locations and have different processing capabilities. The cloud dispatch center adapts the factors according to the comprehensive processing capabilities. Assign audio and video processing tasks, including is the actual processing capacity of the node, The flexibility for nodes to handle different types of tasks, and is the weighting coefficient.

2. The distributed audio and video adaptive production system for multiple terminals according to claim 1, characterized in that: Also includes: Real-time monitoring module: The real-time monitoring module collects the storage utilization of the distributed storage cluster in real time ,in The used storage space, is the total storage space and the task completion rate of the audio and video processing nodes ,in is the number of completed tasks, is the total number of tasks and the energy consumption E of the audio and video processing node; The cloud dispatch center uses data and energy consumption-performance ratio formula Dynamically adjust task allocation and storage strategies.

3. The distributed audio and video adaptive production system for multiple terminals according to claim 1, characterized in that: Also includes: Adaptive coding module: Calculates adaptive coding parameters based on the terminal's network bandwidth B, screen resolution R, processing power P, and content feature vector CFV ,in is the weight coefficient.

4. The distributed audio and video adaptive production system for multiple terminals according to claim 3, characterized in that: The real-time monitoring module also monitors the network delay L between the client and the cloud dispatch center, using the network delay fluctuation coefficient Assess network stability; is the fluctuation range of network delay, is the average network delay; when the network delay fluctuation coefficient LFC exceeds the threshold, the cloud scheduling center adopts an adaptive buffer strategy and adjusts the buffer capacity through the formula Adjust the buffer capacity, where is the cushioning coefficient.

5. The distributed audio and video adaptive production system for multiple terminals according to claim 2, characterized in that: The cloud dispatch center considers the load balancing factor of audio and video processing nodes when adjusting task allocation and the task relevance coefficient; where The processing capacity used by the kth audio and video processing node, is the total processing capacity of the kth audio and video processing node, and n is the number of audio and video processing nodes; the cloud scheduling center maintains a task dependency graph to record the relationship between tasks, and calculates the task correlation coefficient by analyzing the input and output data and processing flow of the tasks; for related tasks, priority is given to nodes that are adjacent or have low communication costs, and the task allocation optimization formula is used. Evaluate allocation options.

6. The distributed audio and video adaptive production system for multiple terminals according to claim 3, characterized in that: The adaptive coding module also considers the complexity C of the audio and video material, and adjusts the adaptive coding parameters based on the complexity C to ,in is the complexity adjustment factor.

7. The distributed audio and video adaptive production system for multiple terminals according to any one of claims 1 to 6, characterized in that: Distributed storage clusters support tiered storage of data based on the access popularity of the data. ,in is the number of data accesses, Calculate the data storage priority for the statistical time period and the timeliness of the data T By storing data on storage media with different performance, and θ are weighting coefficients.

8. The distributed audio and video adaptive production system for multiple terminals according to any one of claims 1 to 6, characterized in that: The client provides audio and video editing functions, supporting users to edit audio and video materials and add subtitles; During the editing process, the system will process the The operation response time estimation formula is calculated based on the operation complexity OC. Optimize operational processes in advance, including is the response time coefficient.

9. The distributed audio and video adaptive production system for multiple terminals according to any one of claims 1 to 6, characterized in that: The system uses encryption algorithms to encrypt audio and video materials during transmission and storage, and uses the encryption strength evaluation formula Evaluate cryptographic security, where is the encryption key length, is the strength coefficient; during the data storage process, the audio and video materials and production projects are encrypted and stored, and the encryption strategy is dynamically adjusted based on the terminal's security level SL. The encryption adjustment formula is .

10. The distributed audio and video adaptive production system for multiple terminals according to any one of claims 1 to 6, characterized in that: The system provides an open API interface, allowing third-party developers to integrate; the success rate formula is called through the API Evaluate API performance, where is the number of successful API calls, is the total number of API calls; If the API call success rate is low, the system will analyze the reasons and predict that it is due to excessive server load or incorrect interface parameters, and make timely optimization and adjustments; the cooperation value index is calculated based on the usage frequency UF and data interaction volume DI of the third-party application. Provide differentiated services and support for different third-party applications, including and is the weighting coefficient.

Images