Autonomous controllable video intelligent compression system, method, equipment and medium

Abstract

The invention relates to the technical field of video processing, and discloses an autonomous controllable intelligent video compression system, method and device and a medium, and the system comprises a routing forwarding unit which is used for carrying out interactive communication with a communication backbone network application service platform, establishing port mapping, and forwarding a control command to a task scheduling unit; the task scheduling unit is used for performing task scheduling with the platform server and the video service terminal, and the task scheduling comprises identity authentication and task distribution; the video intelligent analysis unit is used for performing video processing on the video tasks allocated by the task scheduling unit, wherein the video processing comprises video decoding and video coding; and the video channel selection unit is used for controlling the access of the video terminal. The high-speed routing exchange characteristic is utilized, directional data transmission is achieved, the encoding and decoding functions are achieved through the video intelligent analysis unit, and the video stream data analysis rate is increased; and redundant information is eliminated by using an intelligent compression algorithm, high-fidelity compression transmission is realized, and the robustness of the system is improved.

Description

Technical Field

[0001] The present invention relates to the field of video processing technology, and in particular to an autonomous and controllable video intelligent compression system, method, device and medium. Background Art

[0002] As the construction of new power systems steadily advances, substations, as the heart of power grid operations, face problems such as incomplete operating status of equipment within the station, a surge in the number and types of access equipment, complex transmission backbone networks, and incomplete full-link network monitoring. They need to rely on online intelligent monitoring and ultra-high bandwidth transmission networks to assist in accurate analysis and identification of anomalies, ensure station safety, and improve substation inspection efficiency. Only then can substations achieve digitalization goals, namely, unmanned, centralized, and intelligent operations.

[0003] "Intelligent inspections as the primary method, supplemented by manual inspections" is the current mainstream, economical, and feasible operation and maintenance method for smart substations. Based on the massive amount of data collected, the video content collected, and the real-time frequency of data collection, whether the massive and huge amount of data generated can be processed efficiently, accurately, and quickly directly affects the safe, efficient, and reliable operation of the substation. This requires ultra-large bandwidth to support the transmission network. The investment cost of optical fiber is huge, making it unsuitable for substations with limited budgets. In addition, the bandwidth allocated to video services constrains concurrent transmission performance. In this case, for visual business applications with high bandwidth requirements, the transmission bandwidth can be reduced without affecting the quality of the business content, achieving dozens of times of high-quality and high-rate intelligent compression, that is, lossless compression technology. This can effectively solve the transmission bottlenecks, storage pressure, processing performance, and other problems brought about by massive data. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed. Therefore, the present invention provides an autonomous and controllable video intelligent compression system to solve the problem of improving low-bandwidth and zero-delay transmission of video, ensuring high-definition image quality, and not affecting subsequent intelligent analysis.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] In a first aspect, the present invention provides an autonomous and controllable video intelligent compression system, comprising:

[0007] The routing and forwarding unit is used to interact with the communication backbone network application service platform, establish port mapping, and forward control commands to the task scheduling unit;

[0008] A task scheduling unit is used to perform task scheduling with the platform server and the video service terminal. The task scheduling includes identity authentication and task allocation, and stores and retrieves historical video data.

[0009] a video intelligent analysis unit, configured to perform video processing on the video task assigned by the task scheduling unit, wherein the video processing includes video decoding and video encoding, wherein the video decoding extracts video feature information through a compression algorithm, and the video encoding encodes visual data according to video configuration parameters;

[0010] The video channel selection unit is used to control the access of video terminals and physically connect multiple video terminals to perform adaptive forwarding of video traffic.

[0011] As a preferred solution of the autonomous and controllable video intelligent compression system of the present invention, the workflow of the routing and forwarding unit includes:

[0012] The routing forwarding unit obtains the control command of the platform server and the video stream data monitored by the visual terminal from the communication network;

[0013] The routing forwarding unit sends the video stream data to the task scheduling unit based on the control command.

[0014] As a preferred solution of the autonomous and controllable video intelligent compression system of the present invention, the workflow of the task scheduling unit includes:

[0015] The task scheduling unit communicates interactively with the platform through the routing forwarding unit, parses control commands, and dynamically obtains system parameters;

[0016] The task scheduling unit interacts with the video intelligent analysis unit, assigns video tasks to corresponding idle video intelligent analysis units, and dynamically distributes video compression interaction protocols and compression ratio parameters;

[0017] The beneficial effect of this preferred solution is that the task scheduling unit can achieve efficient task scheduling and management.

[0018] As a preferred solution of the autonomous and controllable video intelligent compression system of the present invention, the workflow of the video intelligent analysis unit includes:

[0019] The video intelligent analysis unit receives the video task and performs task processing. The video intelligent analysis unit is set to multiple. If the video intelligent analysis unit is processing a task, the video intelligent analysis unit is in a busy state. If the video intelligent analysis unit has no task to process, the video intelligent analysis unit is in an idle state.

[0020] The video intelligent analysis unit includes a video acquisition module, a video input analysis module, a video compression algorithm module and a video output reconstruction module, which analyzes, compresses and converts the video data into a format;

[0021] The beneficial effect of this preferred solution is to realize the encoding and decoding function through the video intelligent analysis unit, thereby improving the video stream data analysis rate.

[0022] As a preferred solution of the autonomous and controllable video intelligent compression system of the present invention, the video intelligent analysis unit includes:

[0023] A video acquisition module, configured to acquire compressed video data from the video stream data via a network transmission protocol;

[0024] A video input parsing module, configured to extract video stream service data and decode the video stream from the compressed video acquired by the video acquisition module;

[0025] The video compression algorithm module is used to perform adaptive compression processing on video data through the compression algorithm, eliminate redundant information, and extract video feature information;

[0026] The video output reconstruction module encodes, analyzes, compresses and converts the visual data according to the video configuration parameters, and reconstructs and packages the compressed video stream data for transmission to the next level communication network node according to the multimedia transmission protocol.

[0027] As a preferred solution of the autonomous and controllable video intelligent compression system described in the present invention, the video input parsing module decodes the video, and the business application layer performs underlying decoding operations on video streams in various formats based on the standard streaming media interaction protocol to output complete video source data to the video compression algorithm module.

[0028] As a preferred solution of the autonomous and controllable video intelligent compression system described in the present invention, the video compression algorithm module optimizes video prediction, transformation, motion estimation, and quantization through machine learning and deep learning algorithms, performs redundancy perception processing on the original video data during video compression, automatically identifies the scene and target of the video screen, and adaptively adjusts the encoding parameters to perform lossless compression;

[0029] The beneficial effect of this preferred solution is that it utilizes an intelligent compression algorithm to eliminate redundant information and achieve high-fidelity compressed transmission.

[0030] In a second aspect, the present invention provides an autonomous and controllable video intelligent compression method, comprising:

[0031] The routing forwarding unit receives the control command issued by the service platform and forwards it to the task scheduling unit through the port mapping relationship;

[0032] The task scheduling unit analyzes the control signaling, coordinates the idle video intelligent analysis units in the local area network, issues task instructions, and performs protocol forwarding of control commands;

[0033] Based on the task requirements of the control command, the video intelligent analysis unit establishes a point-to-point streaming media communication link with the video service terminal to obtain video source data;

[0034] The video source data is decoded by the video input parsing module of the video intelligent analysis unit, the decoded video data is hierarchically processed by the video compression algorithm module, the compression algorithm is integrated, and redundancy perception, elimination and reconstruction of the video information stream are performed, and the compressed video data stream is reconstructed and encapsulated by the video output reconstruction module;

[0035] The generated standard stream video data is uploaded to the service platform or forwarded to the application service platform in the local area network through the routing forwarding unit or the video channel selection unit according to the network topology deployed by the main station.

[0036] In a third aspect, the present invention provides a computer device, comprising:

[0037] memory and processor;

[0038] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the autonomous and controllable video intelligent compression system are implemented.

[0039] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the autonomous and controllable video intelligent compression system.

[0040] Compared with the existing technology, the present invention has the following beneficial effects: the present invention utilizes the characteristics of high-speed routing and switching to realize directional data transmission and strengthen video access control; realizes encoding and decoding functions through the video intelligent analysis unit, and improves the video stream data parsing rate; utilizes the intelligent compression algorithm to eliminate redundant information and high-fidelity compression transmission, thereby improving the robustness of the system; not only can the video compression rate be dynamically adjusted to save storage costs, but also bypass access to existing video access terminals can be achieved, thereby realizing safe and stable operation of substation video services. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0042] Figure 1 This is a logical function diagram of a video compression gateway of an autonomous and controllable video intelligent compression system according to an embodiment of the present invention;

[0043] Figure 2 This is a flow chart of intelligent compression encoding of an autonomous and controllable video intelligent compression system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0045] Reference Figure 1-Figure 2 , as an embodiment of the present invention, provides an autonomous and controllable video intelligent compression system, comprising:

[0046] The routing and forwarding unit is used to interact with the communication backbone network application service platform, establish port mapping, forward control commands to the task scheduling unit, and provide a unified communication interface for uploading surveillance videos;

[0047] The task scheduling unit is used to schedule tasks between the platform server and the video service terminal. Task scheduling includes identity authentication and task allocation, and stores and retrieves historical video data. It supports large-capacity hard disk storage expansion and has the function of local storage and retrieval of historical video data.

[0048] The video intelligent analysis unit is used to perform video processing on the video tasks assigned by the task scheduling unit. Video processing includes video decoding and video encoding. Video decoding extracts video feature information through an intelligent compression algorithm, and video encoding encodes visual data according to video configuration parameters.

[0049] The video channel selection unit is used to control the access of video terminals and physically connect multiple video terminals to perform adaptive forwarding of video traffic.

[0050] It should be noted that the autonomous and controllable video intelligent compression system proposed in this invention is suitable for intelligent inspection applications in substations. The core is the video super intelligent compression gateway. Under the premise of ensuring the quality of video data, it achieves high-quality and high-rate compression of dozens of times, which to a certain extent alleviates the practical difficulties such as transmission bottlenecks, storage pressure, and processing performance brought about by massive video data faced by monitoring applications. The system hardware structure diagram is shown in the figure. Figure 1 As shown, it mainly includes a routing forwarding unit, a task scheduling unit, a video intelligent analysis unit, and a video channel selection unit, wherein a maximum of 8 video intelligent analysis units are supported to run concurrently, that is, N is 8.

[0051] In a preferred embodiment, the workflow of the routing and forwarding unit includes:

[0052] The routing and forwarding unit obtains the control commands of the platform server and the video stream data monitored by the visual terminal from the communication network;

[0053] The routing and forwarding unit sends the video stream data to the task scheduling unit based on the control command.

[0054] In an optional embodiment, the routing forwarding unit is composed of a network interface module and a central processing unit chip, the network interface module includes an adaptive Ethernet physical layer chip, a network transformer and an interface device, and the central processing unit chip is provided with a data link layer Gigabit MAC, a wireless communication interface, an Ethernet configuration port and a business data communication interface, wherein the Ethernet configuration port is provided with a localized port forwarding application, and the business data interface is provided with a two-way communication interaction channel to transmit video compressed data.

[0055] In another optional embodiment, the routing forwarding unit is composed of a network interface module and a Hi5651T Lingxiao quad-core CPU chip. The network interface module refers to an adaptive 10 / 100 / 1000Mbps Ethernet physical layer PHY chip, a network transformer and an interface device. The Hi5651T chip is provided with a data link layer Gigabit MAC and WiFi wireless communication interface, providing an Ethernet configuration port and a service data communication interface, wherein the configuration port provides a localized port forwarding application, and the service data interface provides a two-way communication interaction channel for transmitting video compressed data.

[0056] In a preferred embodiment, the workflow of the task scheduling unit includes:

[0057] The task scheduling unit communicates with the platform through the routing forwarding unit, parses control commands, and dynamically obtains system parameters;

[0058] The task scheduling unit interacts with the video intelligent analysis unit, assigns video tasks to the corresponding idle video intelligent analysis units, and dynamically distributes video compression interaction protocols and compression ratio parameters.

[0059] In an optional embodiment, the task scheduling unit is designed for the RK3568J platform. It communicates northbound with the platform through the routing and forwarding unit, interprets control commands, and dynamically obtains key system parameters. It also interacts horizontally with the video intelligent analysis unit to dynamically distribute video compression interaction protocols and compression ratio parameters. Key system parameters primarily refer to those that have a significant impact on the operation and performance of the entire autonomous and controllable video intelligent compression system. These can be network-related parameters, video-related parameters, etc. The compression ratio parameter can be set according to the video compression interaction protocol, or according to a preset compression ratio value, based on the actual application scenario or video format.

[0060] It should be noted that the task scheduling unit can fully utilize the processing power, storage capacity and communication capacity of RK3568J to achieve efficient task scheduling and management. The video compression interaction protocol stipulates the data interaction method and rules between the video intelligent analysis unit and other modules when performing video processing; the compression ratio parameter determines the degree of compression of video data during the encoding process. Different compression ratios will affect the video quality and file size.

[0061] In another optional implementation, the task scheduling unit can also use distributed scheduling and a microservice architecture to decompose task scheduling into independent services such as identity authentication, parameter management, and data routing. It supports dynamic scaling and adapts to large-scale video terminal access. A lightweight scheduling module is deployed at the edge to be responsible for real-time task allocation of local video terminals; the center performs global resource coordination (such as cross-regional video stream scheduling and computing power load balancing); and integrates machine learning models to dynamically optimize scheduling strategies based on historical task data, network status, device load, etc., such as automatically adjusting compression ratio distribution and prioritizing high-priority video streams.

[0062] In a preferred embodiment, the workflow of the video intelligent analysis unit includes:

[0063] The video intelligent analysis unit receives video tasks and processes them. If there are multiple video intelligent analysis units, the video intelligent analysis unit is in a busy state if the video intelligent analysis unit is processing a task. If the video intelligent analysis unit has no task to process, the video intelligent analysis unit is in an idle state.

[0064] The video intelligent analysis unit includes a video acquisition module, a video input analysis module, a video compression algorithm module and a video output reconstruction module, which analyzes, compresses and converts the video data into a format.

[0065] In a preferred embodiment, the video intelligent analysis unit includes:

[0066] High-definition video acquisition module, used to obtain compressed video data from video stream data through network transmission protocol;

[0067] The video input parsing module is used to extract video stream service data and decode the video stream from the compressed video acquired by the high-definition video acquisition module;

[0068] The video compression algorithm module is used to perform adaptive compression processing on video data through the compression algorithm, eliminate redundant information, and extract video feature information;

[0069] The video output reconstruction module encodes, analyzes, compresses and converts the visual data according to the video configuration parameters, and reconstructs and packages the compressed video stream data according to the multimedia transmission protocol and transmits it to the next-level communication network node, which can minimize transmission delay.

[0070] In an optional implementation, the high-definition video acquisition module acquires compressed high-definition video data from the camera via the Real Time Streaming Protocol (RTSP), and the format of the high-definition video data is H.264 or H.265 encoding.

[0071] In a preferred embodiment, the video input parsing module performs decoding based on the main control hardware, and the business application layer performs underlying decoding operations on video streams in various formats based on the standard streaming media interaction protocol, outputting complete video source data for the compression processing operation of the video compression algorithm module.

[0072] In an optional embodiment, the video input parsing module is used to extract video stream service data and decode the compressed video obtained by the high-definition video acquisition module, supporting hardware encoding and decoding and 6TOPS computing power. 6TOPS computing power means that the processor can perform 6 trillion basic operations per second. It is a key indicator for measuring the computing power of computing devices (such as graphics cards, AI chips, etc.). TOPS (Tera Operations Per Second) is the unit of processor computing power. 1TOPS corresponds to 1 trillion operations per second (such as addition, multiplication, etc.). The video input parsing module is based on the host RK3576 main control hardware decoding function, the business application layer is based on the standard streaming media interaction protocol, and the underlying decoding operation supports commonly used H.264 / H.265 and other format video streams, thereby providing complete video source data for compression processing operations.

[0073] In a preferred embodiment, the video compression algorithm module optimizes video prediction, transformation, motion estimation, and quantization through machine learning and deep learning algorithms, performs redundant perception processing on the original video data during video compression, automatically identifies the scenes and targets of the video screen, adaptively adjusts the encoding parameters, and performs lossless compression.

[0074] In an optional embodiment, a video compression model is constructed by a compression algorithm. Figure 2 This is a flowchart of the intelligent compression coding of the autonomous and controllable video intelligent compression system of the present invention. By integrating advanced machine learning and deep learning technologies, the present invention has made unique and important improvements in prediction (intra-frame and inter-frame), transformation, quantization, etc., including:

[0075] Step 1 is based on perceptual refinement coding, which selectively adjusts the way different areas are divided into blocks. For example, the image in the target area is forcibly divided into 8×8 or 4×4 fine units, a fine prediction model is established, and more codewords are used to describe these image areas to achieve better target area image quality. For images in the background area, the prediction unit division accuracy can be reduced, for example, to 32×32 or 16×16, and some background image details and update frequency can be moderately ignored, effectively reducing the bit rate of the background area image.

[0076] It should be noted that intelligent compression technology combines two unique and complementary solutions: one is salient object detection for general videos, and the other is image semantic segmentation for specific scenes. The characteristics of these two solutions are: salient object detection is modeled based on people's attention to the picture, which can minimize the impact on image quality while reducing the bit rate of non-salient areas; while semantic segmentation can mark different scenes in the picture and dynamically adjust the encoding parameters, such as pedestrians, faces, vehicles, signs, etc.

[0077] Step 2: A deep convolutional neural network is used to optimize fast intra-frame mode selection and improve intra-frame prediction efficiency. Prediction units and corresponding prediction modes extracted from massive surveillance videos are used for training and applied in the encoding process. First, a classifier is selected based on the size of the prediction unit. The pixels of the prediction unit are then fed into the network, and one of 35 modes is output as the intra-frame prediction mode for the current prediction unit.

[0078] In addition, the "fast intra-frame prediction mode selection method" of intelligent compression technology adopts online learning and regularly uses a general intra-frame prediction method to correct the network structure and parameters to reduce the cumulative prediction error and adapt to the ever-changing coding video environment.

[0079] Step three: intelligent optimization of quantization parameters. In the encoding process, quantization is the only step that loses data accuracy and introduces errors. The video frame is divided into target content, and a smaller quantization parameter value is used in the target area to retain as many image details as possible in the target area, while a larger quantization parameter value is used in the background area to achieve efficient compression of the overall video bit rate.

[0080] It should be noted that the video compression model of the present invention can strike a balance between coding efficiency and computational complexity. That is, it is based on the efficient hardware encoding and decoding functions of Rockchip RK3576 and the model training implemented by deep compression software. The efficient division of labor between software and hardware jointly reduces the implementation requirements of intelligent compression technology.

[0081] In another optional implementation, the video compression model constructed by the compression algorithm can also be constructed through more than 60 advanced machine learning and deep learning technologies such as target detection, target tracking, semantic segmentation, structural point detection, knowledge distillation, and compressed sensing. It can perform intelligent optimization in prediction (intra-frame, inter-frame), transformation, motion estimation, quantization, etc., perform redundant perception processing on the original video data during compression, automatically identify the scenes and targets of the video screen, and adaptively adjust the encoding parameters to achieve lossless compression.

[0082] In an optional implementation, the video output reconstruction module re-encodes and protocol packages the compressed video data according to the configuration parameters to form a standard video stream that complies with Ethernet transmission.

[0083] In an optional implementation, the video channel selection unit is composed of a dedicated switching chip, adopts a modular design, and is adapted according to the number of local video surveillance terminals required to access the system.

[0084] In an optional embodiment, the video channel selection unit is composed of a dedicated switching chip, and the hardware selection may include Hi5651T or Kun Gao. It adopts a modular design and can be adapted to plug and play according to the number of local video surveillance terminal access requirements. Among them, the Hi5651T platform routing solution is recommended for video terminal access to gigabit transmission bandwidth. If the bandwidth requirement is higher than gigabit, the KG6524 hardware solution can be selected, which can meet the deployment and operation and maintenance applications of the in-station LAN intelligent patrol monitoring platform.

[0085] It should be noted that the present invention utilizes high-speed routing and switching characteristics to achieve directional data transmission and strengthen video access control; implements encoding and decoding functions through a video intelligent analysis unit to improve the video stream data parsing rate; utilizes an intelligent compression algorithm to eliminate redundant information and achieve high-fidelity compression transmission, thereby improving the robustness of the system; not only can the video compression rate be dynamically adjusted to save storage costs, but also bypass access to existing video access terminals to achieve safe and stable operation of substation video services.

[0086] Example 2

[0087] This embodiment provides an autonomous and controllable video intelligent compression method, including:

[0088] The routing forwarding unit receives the control command issued by the service platform and forwards it to the task scheduling unit through the port mapping relationship;

[0089] The task scheduling unit analyzes control signals, coordinates idle video intelligent analysis units in the local area network, issues task instructions, and forwards control commands according to the protocol;

[0090] Based on the task requirements of the control command, the video intelligent analysis unit establishes a point-to-point streaming communication link with the video service terminal to obtain video source data;

[0091] The video input parsing module of the video intelligent analysis unit decodes the video source data, and the video compression algorithm module performs hierarchical processing on the decoded video data, integrates the compression algorithm, and performs redundant perception, elimination, and reconstruction of the video information stream. The video output reconstruction module reconstructs and encapsulates the compressed video data stream;

[0092] Through the routing forwarding unit or the video channel selection unit, the generated standard stream video data is uploaded to the service platform or forwarded to the application service platform within the local area network according to the network topology deployed by the main station.

[0093] The above is a schematic diagram of an autonomous and controllable video intelligent compression system according to this embodiment. It should be noted that the technical solution of this autonomous and controllable video intelligent compression method and the technical solution of the autonomous and controllable video intelligent compression system described above share the same concept. For details not described in detail in the technical solution of the autonomous and controllable video intelligent compression method according to this embodiment, please refer to the description of the technical solution of the autonomous and controllable video intelligent compression system described above.

[0094] This embodiment further provides a computer device suitable for autonomous and controllable video intelligent compression, including:

[0095] Memory and processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the autonomous and controllable video intelligent compression system proposed in the above embodiment.

[0096] This embodiment further provides a storage medium storing a computer program, which, when executed by a processor, implements the autonomous and controllable video intelligent compression system proposed in the above embodiment.

[0097] The storage medium proposed in this embodiment and the autonomous and controllable video intelligent compression system proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0098] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general-purpose hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a computer's floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0099] Example 3

[0100] Referring to Table 1, an embodiment of the present invention provides an autonomous and controllable video intelligent compression system. In order to verify its beneficial effects, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0101] The system of the present invention has excellent performance in video compression. A single video intelligent analysis unit supports online compression of 16-channel H.264 / H.265 format video streams with a maximum resolution of 1080P. Through advanced machine vision, artificial intelligence retrieval and other technologies, it can perform an average of 10 times compression without changing the frame rate, resolution and duration of the original video. The specific compression performance parameters are shown in Table 1, with a resolution of 1080P as the basic condition.

[0102] Table 1: Video intelligent compression parameter values

[0103] Input encoding format Output encoding Average compression ratio Video source description H.264 H.264 50％ Mixed static and dynamic scenes H.264 H.265 90％ Mixed static and dynamic scenes H.265 H.265 50％ Mixed static and dynamic scenes

[0104] The present invention uses artificial intelligence technology to compress video data. It can provide real-time video compression without affecting the original resolution, video frame rate, and video duration, reducing data redundancy and achieving a significant compression ratio. After compressing the surveillance images in a visually lossless manner, the storage capacity can be saved by an average of about 60%. It is compatible with H.264 / H.265 mainstream encoding technologies and also supports autonomous encoding applications. It outputs to mainstream protocols such as standard RTSP, reducing the bit rate by up to 90% compared to the original video, alleviating the pressure on the concurrent transmission bandwidth of network video streams. At the same time, under the premise of equal storage capacity, it greatly extends the video storage period and enhances the diversity of video surveillance application solutions.

[0105] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An autonomous and controllable video intelligent compression system, characterized in that: include: The routing and forwarding unit is used to interact with the communication backbone network application service platform, establish port mapping, and forward control commands to the task scheduling unit; A task scheduling unit is used to perform task scheduling with the platform server and the video service terminal. The task scheduling includes identity authentication and task allocation, and stores and retrieves historical video data. a video intelligent analysis unit, configured to perform video processing on the video task assigned by the task scheduling unit, wherein the video processing includes video decoding and video encoding, wherein the video decoding extracts video feature information through a compression algorithm, and the video encoding encodes visual data according to video configuration parameters; The video channel selection unit is used to control the access of video terminals and physically connect multiple video terminals to perform adaptive forwarding of video traffic.

2. The autonomous and controllable video intelligent compression system according to claim 1, characterized in that: The workflow of the routing forwarding unit includes: The routing forwarding unit obtains the control command of the platform server and the video stream data monitored by the visual terminal from the communication network; The routing forwarding unit sends the video stream data to the task scheduling unit based on the control command.

3. The autonomous and controllable video intelligent compression system according to claim 1 or 2, characterized in that: The workflow of the task scheduling unit includes: The task scheduling unit communicates interactively with the platform through the routing forwarding unit, parses control commands, and dynamically obtains system parameters; The task scheduling unit interacts with the video intelligent analysis unit, allocates video tasks to corresponding idle video intelligent analysis units, and dynamically distributes video compression interaction protocols and compression ratio parameters.

4. The autonomous and controllable video intelligent compression system according to claim 1, wherein: The workflow of the video intelligent analysis unit includes: The video intelligent analysis unit receives the video task and performs task processing. The video intelligent analysis unit is set to multiple. If the video intelligent analysis unit is processing a task, the video intelligent analysis unit is in a busy state. If the video intelligent analysis unit has no task to process, the video intelligent analysis unit is in an idle state. The video intelligent analysis unit includes a video acquisition module, a video input analysis module, a video compression algorithm module and a video output reconstruction module, which analyzes, compresses and converts the video data into a format.

5. The autonomous and controllable video intelligent compression system according to claim 4, characterized in that: The video intelligent analysis unit includes: A video acquisition module, configured to acquire compressed video data from the video stream data via a network transmission protocol; A video input parsing module, configured to extract video stream service data and decode the video stream from the compressed video acquired by the video acquisition module; The video compression algorithm module is used to perform adaptive compression processing on video data through the compression algorithm, eliminate redundant information, and extract video feature information; The video output reconstruction module encodes, analyzes, compresses and converts the visual data according to the video configuration parameters, and reconstructs and packages the compressed video stream data for transmission to the next level communication network node according to the multimedia transmission protocol.

6. The autonomous and controllable video intelligent compression system according to claim 5, characterized in that: The video input parsing module decodes the video, and the service application layer performs underlying decoding operations on video streams in various formats based on the standard streaming media interaction protocol to output complete video source data to the video compression algorithm module.

7. The autonomous and controllable video intelligent compression system according to claim 6, characterized in that: The video compression algorithm module optimizes video prediction, transformation, motion estimation, and quantization through machine learning and deep learning algorithms, performs redundant perception processing on the original video data during video compression, automatically identifies the scenes and targets of the video screen, adaptively adjusts encoding parameters, and performs lossless compression.

8. An autonomous and controllable video intelligent compression method, using an autonomous and controllable video intelligent compression system according to any one of claims 1 to 7, characterized in that: include, The routing forwarding unit receives the control command issued by the service platform and forwards it to the task scheduling unit through the port mapping relationship; The task scheduling unit analyzes the control signaling, coordinates the idle video intelligent analysis units in the local area network, issues task instructions, and performs protocol forwarding of control commands; Based on the task requirements of the control command, the video intelligent analysis unit establishes a point-to-point streaming media communication link with the video service terminal to obtain video source data; The video source data is decoded by the video input parsing module of the video intelligent analysis unit, the decoded video data is hierarchically processed by the video compression algorithm module, the compression algorithm is integrated, and redundancy perception, elimination and reconstruction of the video information stream are performed, and the compressed video data stream is reconstructed and encapsulated by the video output reconstruction module; The generated standard stream video data is uploaded to the service platform or forwarded to the application service platform in the local area network through the routing forwarding unit or the video channel selection unit according to the network topology deployed by the main station.

9. A computer device, characterized in that: include: memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, an autonomous and controllable video intelligent compression system as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that It stores computer-executable instructions, which, when executed by a processor, can implement an autonomous and controllable video intelligent compression system as described in any one of claims 1 to 7.

Images