Algorithm collaboration system based on comprehensive video platform

Abstract

The invention is suitable for the technical field of data, and provides an algorithm collaboration system based on a comprehensive video platform, which comprises a comprehensive video platform, an algorithm collaboration platform and a camera cluster, the comprehensive video platform is used for acquiring picture sample data from the camera cluster and providing a data interface for the algorithm collaboration platform; the algorithm collaboration platform comprises an algorithm integration module which is used for providing a preset algorithm for an algorithm verification module; the algorithm verification module is used for calling a preset algorithm to carry out target identification on the picture sample data to obtain a research and judgment result of the picture sample data; the algorithm collaboration platform sends a research and judgment result corresponding to the picture sample data to the comprehensive video platform through the data interface; and the camera cluster is used for intercepting video data according to an acquisition task of the comprehensive video platform to obtain picture sample data. According to the algorithm collaboration system based on the comprehensive video platform, collaboration processing of video data and algorithms is realized, and the algorithm collaboration system based on the comprehensive video platform has practicability and usability.

Description

Technical Field

[0001] The present application belongs to the field of data processing technology, and in particular, relates to an algorithm collaboration system based on a comprehensive video platform. Background Art

[0002] With the continuous development of video surveillance technology, integrated video platforms have become important infrastructure in the fields of smart cities and smart transportation. However, with the explosive growth of video data and the increasing application of algorithms, how to achieve algorithm collaboration and efficient management has become a major challenge facing the current integrated video platform.

[0003] Most existing comprehensive video platforms focus on the collection, storage and playback of video data, and lack a unified management and coordination mechanism for AI algorithms. This leads to inefficient algorithm application and difficulty in meeting the needs of different scenarios. In addition, the algorithm data management and analysis capabilities are insufficient, making it difficult to conduct effective algorithm evaluation and optimization, which restricts the implementation of the algorithm. Summary of the invention

[0004] The embodiment of the present application provides an algorithm collaboration system based on a comprehensive video platform, which can solve the problem that video data and AI algorithms on the existing comprehensive video platform are difficult to achieve collaborative processing.

[0005] In a first aspect, an embodiment of the present application provides an algorithm collaboration system based on a comprehensive video platform, including:

[0006] The integrated video platform is used to obtain sample data from the camera cluster and provide a data interface to the algorithm collaboration platform;

[0007] The algorithm collaboration platform comprises an algorithm integration module and an algorithm verification module, wherein the algorithm integration module is used to provide a preset algorithm to the algorithm verification module; and the algorithm verification module is used to obtain the picture sample data from the data interface, call the preset algorithm to perform target recognition on the picture sample data, and obtain the analysis result of the picture sample data; the algorithm collaboration platform sends the analysis result corresponding to the picture sample data to the comprehensive video platform through the data interface;

[0008] The camera cluster is used to acquire video data, and intercept the video data according to the acquisition task of the integrated video platform to obtain the picture sample data, wherein the picture sample data includes picture data and / or real-time data stream.

[0009] Compared with the prior art, the embodiments of the present application have the following beneficial effects: the algorithm collaboration system based on the comprehensive video platform of the present application includes a comprehensive video platform, an algorithm collaboration platform and a camera cluster; the comprehensive video platform is used to obtain picture sample data from the camera cluster, and provide a data interface to the algorithm collaboration platform; the algorithm collaboration platform includes an algorithm integration module for providing a preset algorithm to the algorithm verification module; and the algorithm verification module is used to call the preset algorithm to perform target recognition on the picture sample data to obtain the judgment result of the picture sample data; the algorithm collaboration platform sends the judgment result corresponding to the picture sample data to the comprehensive video platform through the data interface; the camera cluster is used to intercept the video data according to the acquisition task of the comprehensive video platform to obtain the picture sample data. The present application realizes the collaboration between video data and algorithm through the collaborative processing of the comprehensive video platform, the algorithm collaboration platform and the camera cluster, and the present application is practical and easy to use. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0011] Figure 1 It is a schematic diagram of the overall architecture of an algorithm collaboration system based on a comprehensive video platform provided in one embodiment of the present application;

[0012] Figure 2 It is a schematic diagram of the workflow of an algorithm collaboration system based on a comprehensive video platform provided in one embodiment of the present application. DETAILED DESCRIPTION

[0013] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0014] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0015] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0016] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0017] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0018] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0019] Figure 1This is the first schematic diagram of an algorithm collaboration system based on a comprehensive video platform provided by an embodiment of the present application, and the algorithm collaboration system based on the comprehensive video platform includes a comprehensive video platform, an algorithm collaboration platform and a camera cluster. Among them, the comprehensive video platform (i.e., the comprehensive video management platform) is used to obtain picture sample data from the camera cluster, and to provide a data interface for data interaction between the comprehensive video platform and the algorithm collaboration platform. The algorithm collaboration platform (i.e., the AI ​​algorithm collaborative innovation platform) includes an algorithm integration module and an algorithm verification module, wherein the algorithm integration module is used to provide a preset algorithm to the algorithm verification module; and the algorithm verification module is used to obtain picture sample data from the data interface, call the preset algorithm to perform target recognition on the picture sample data, and obtain the judgment result of the picture sample data; the algorithm collaboration platform sends the judgment result corresponding to the picture sample data to the comprehensive video platform through the data interface. The camera cluster is used to obtain video data, and to intercept the video data according to the acquisition task of the comprehensive video platform to obtain picture sample data, and the picture sample data includes picture data and / or real-time data stream.

[0020] Exemplarily, the camera cluster includes more than four camera devices (ie, Camera).

[0021] Specifically, the algorithm verification module of the algorithm collaboration platform can access and aggregate video data from camera clusters, integrated video platforms, etc., and call on multiple established target recognition algorithms to capture and identify specific scene images respectively, and store and count recognition results, sample images and other data in a register.

[0022] The algorithm collaboration system based on the comprehensive video platform realizes efficient video data processing and analysis by integrating the comprehensive video platform, the algorithm collaboration platform and the camera cluster. The system improves the accuracy and speed of target recognition by integrating multiple algorithms, and realizes real-time analysis and judgment of the image sample data captured by the camera cluster. At the same time, the data interface of the comprehensive video platform is used to realize the data interaction between the comprehensive video platform and the algorithm collaboration platform for the image sample data, ensuring the flexible deployment and optimization of the algorithm, thereby realizing intelligent recognition and early warning in video surveillance scenarios, and improving the intelligence level and emergency response capabilities of the overall video surveillance system.

[0023] In some embodiments, the comprehensive video platform includes a storage module, which is used to store picture sample data and analysis results. Specifically, the storage module supports the storage of picture sample data such as pictures locally, and the local storage is on the local server hard disk, disk array (such as FC-SAN, IP-SAN, etc.). Exemplarily, the analysis results include good or poor; if the analysis result is poor and the picture sample data is in picture format data, the storage period of the picture sample data is set to be greater than 180 days. That is, the storage period of the alarm picture is not less than 180 days. The storage module supports structured data and resource file storage, where picture files should support mainstream formats such as JPEG, BMP, PNG, and video files should support general formats such as mp4, avi, and custom encrypted storage formats. The storage module provides index management, data retrieval, and data retrieval services for stored data. The storage module supports data integrity and health status checks, data disaster recovery backup and recovery.

[0024] By building a collaborative system of comprehensive video platform, algorithm collaboration platform and camera cluster, data interaction between platforms is realized through data interface, and the cooperation of algorithm integration module and algorithm verification module is realized to achieve efficient target recognition and judgment of image sample data, realize intelligent analysis and processing of video data, improve the automation level and response speed of video surveillance system, and the integration of storage module ensures the persistent preservation of data and historical backtracking capability.

[0025] Specifically, the integrated video platform includes an access module to support access to video terminal devices through various types and versions of protocols such as GB / T28181, ONVIF, and mainstream terminal manufacturers. The integrated video platform can support access to cameras from multiple manufacturers, IP video encoding devices such as NVR, and IP video encoding devices from mainstream manufacturers. The integrated video platform can support the connection of various video platforms in accordance with the GB / T28181 protocol. The integrated video platform can connect to the established target recognition algorithm API.

[0026] Specifically, the integrated video platform includes a streaming media streaming module, which is used to support standard bitstreams through RTSP, RTMP, and HLS protocols, FLV forwarding (http-flv, websocket-flv), and supports real-time streaming through URLs. The integrated video platform includes a view decoding module, which is used to parse and capture standard bitstreams into images for downstream applications to call or store.

[0027] Specifically, the comprehensive video platform includes a live video processing module, which is used to support real-time video browsing of authorized monitoring points on a PC via the web; for picture capture, the system supports timed capture based on fixed preset positions; users can capture the required images when watching real-time video or recording; for pan-tilt control, users can remotely control the rotation, zoom, magnification, aperture and other functions of the pan-tilt camera; for supporting video playback in the real-time monitoring window; for fuzzy query, it supports fuzzy query function of front-end monitoring points, and the camera that needs to be viewed can be quickly found through the camera name keywords.

[0028] Specifically, the integrated video platform includes a video recording and playback calling module, which is used to call the recording function of the existing integrated video management platform, and supports video query and playback according to the recording date and time. In addition, the video recording and playback calling module is used to call the recording control function of the existing integrated video management platform, and supports video playback control: supporting timeline playback of videos; supporting timeline support for dragging and clicking operations to quickly locate videos; supporting multi-channel synchronous playback operations; supporting pause and multiple fast and slow playback.

[0029] Furthermore, the algorithm collaboration platform also includes an algorithm configuration module, which is used to configure a preset algorithm according to a preset scene; the camera cluster includes a first control module and more than four first cameras; and the storage module also stores a first association table, which is used to save the first association between the preset algorithm information and the camera information. After the preset algorithm configuration is completed, the algorithm collaboration platform sends configuration information to the comprehensive video platform, and the configuration information includes the first preset algorithm information. The comprehensive video platform queries the first association table according to the configuration information to determine the first camera information that has a first association with the first preset algorithm information; the comprehensive video platform sends the first acquisition task and the first camera information to the camera cluster. The first control module is used to control the first camera corresponding to the first camera information, intercept the video data according to the first acquisition task, and obtain the first screen sample data; the first screen sample data is used as the screen sample data.

[0030] Exemplarily, the preset algorithm includes one of a deep learning-based target detection algorithm, a single-stage target detection algorithm, a key point-based target detection algorithm, and a Transformer-based target detection algorithm.

[0031] Specifically, the scenarios in which the established target recognition algorithm (i.e., the preset algorithm) is applicable are different. The algorithm collaboration platform should provide the preset algorithm for binding with the cameras in the camera cluster (i.e., the first association relationship between the preset algorithm and the camera device information), that is: dynamically assigning one or more cameras (i.e., camera devices) in the camera cluster to a specific algorithm (i.e., the preset algorithm) to participate in the calculation and verification of the algorithm.

[0032] By integrating the collaborative work of the comprehensive video platform, the algorithm collaboration platform and the camera cluster, configuring the preset algorithm according to the preset scene through the algorithm configuration module, and realizing effective data interaction between platforms through the data interface, intelligent video analysis for specific scenes is realized; by establishing a first association relationship table to store the association between algorithm information and camera equipment information, precise control and task allocation of the camera cluster are realized; through the cooperation of the control module and the camera equipment, video data is intercepted according to the collection task, efficient video sample collection and target recognition are realized, thereby realizing the intelligent, automated and precise management of the video surveillance system.

[0033] In some embodiments, the algorithm collaboration platform also includes an algorithm scheduling module, which is used to call the first data interface corresponding to the preset algorithm of the integrated video platform, and use the first data interface as the aforementioned data interface. Furthermore, the algorithm scheduling module calls the algorithm verification module to receive the picture sample data from the integrated video platform through the first data interface, wherein the picture sample data is a real-time data stream. And, after obtaining the judgment result of the picture sample data, the algorithm scheduling module determines whether the judgment result is good; if not, the algorithm scheduling module generates alarm event data and stores it in the database of the algorithm collaboration platform, and the alarm event data includes alarm contents such as timestamp, picture sample data, and judgment result.

[0034] Furthermore, the algorithm collaboration platform provides an interactive interface. After receiving a click command, the interactive interface switches to the alarm event details page. The alarm event details page displays the alarm event data, and the alarm event data includes alarm content, longitude and latitude, alarm type, etc.

[0035] In some embodiments, the algorithm integration module stores service specifications corresponding to the preset algorithm; the algorithm collaboration platform also includes an algorithm docking module, which is used to configure the preset algorithm according to the service specifications so that the preset algorithm is adapted to the algorithm verification module.

[0036] Specifically, the algorithm verification module is also used to support manual verification and judgment of the recognition results of the preset algorithm; to support the function of recording target images or videos that are missed by the preset algorithm; and to support statistics and export of sample image data related to the recognition of the preset algorithm.

[0037] Through the algorithm collaboration platform that integrates the algorithm integration module and the algorithm verification module, by storing the service specifications corresponding to the preset algorithm, and configuring the preset algorithm according to the service specifications through the algorithm docking module, the adaptation and efficient docking of the preset algorithm and the algorithm verification module are realized, thereby realizing the rapid deployment and precise application of the algorithm; through the interactive capability of the data interface, the data flow between the comprehensive video platform and the algorithm collaboration platform and the real-time feedback of the analysis and judgment results are realized, thereby improving the intelligent analysis and processing capabilities of the video surveillance system.

[0038] Furthermore, the camera cluster includes a second control module and more than four second camera devices; the algorithm collaboration platform also includes a scene point selection module and a point capture configuration module, wherein the scene point selection module is used to determine the first algorithm type according to the preset algorithm after the algorithm docking module configures the preset algorithm according to the service specification. The scene point selection module stores a second association relationship table, and the second association relationship table is used to save the second association relationship between the algorithm type and the point information and the scene information. The scene point selection module queries the second association relationship table according to the first algorithm type to determine the first point information and the first scene information that have a second association relationship with the first algorithm type. The point capture configuration module is used to determine the preset position of the second camera device according to the first point information and the first scene information. The second control module is used to control the second camera device to move to the preset position, intercept the video data according to the second acquisition task, and obtain the second screen sample data; the second screen sample data is used as the screen sample data.

[0039] By constructing an algorithm collaboration platform including a scene point selection module and a point capture configuration module, the first algorithm type is determined after the preset algorithm is configured according to the service specification in the algorithm docking module, and the association between the algorithm type and the point information and the scene information is stored through the second association table, so as to achieve precise matching of the algorithm type with the points of a specific monitoring scene; the preset position of the camera device is determined through the point capture configuration module, and the second control module controls the camera device to move to the preset position and execute the acquisition task, so as to achieve refined video data acquisition for specific scenes; through these technical means, the highly automated, intelligent and data-driven target recognition and analysis of the video surveillance system is achieved, thereby improving the overall monitoring efficiency and response speed of the system.

[0040] Furthermore, the algorithm collaboration platform also includes a startup module, a detection and analysis module and a detection statistics module, wherein the startup module is used to call the preset algorithm to perform target recognition on the picture sample data after the algorithm docking module configures the preset algorithm according to the service specification to obtain the recognition result of the picture sample data. The detection and analysis module is used to judge the recognition result according to the preset analysis and judgment rules to obtain the analysis and judgment result. The statistical display module is used to visualize the recognition result and the analysis and judgment result corresponding to the recognition result within a unit time.

[0041] Specifically, the startup module is used to support custom strategies such as timing and quantity to automatically schedule camera clusters to obtain video data and send it into the preset algorithm for target recognition and verification; and to support compatibility and adaptation with applicable requirements of other business departments, that is, not interfering with the normal operation of platforms such as cameras and video surveillance.

[0042] Through the algorithm collaboration platform integrating the startup module, detection and analysis module and detection statistics module, after configuring the preset algorithm according to the service specification in the algorithm docking module, the preset algorithm is called to perform target recognition on the image sample data and obtain the recognition result, and the recognition result is judged by the detection and analysis module according to the preset analysis and judgment rules to realize the intelligent analysis and judgment of the video surveillance data; the recognition result and the analysis result are visualized in unit time through the statistical display module to realize real-time statistics and intuitive display of monitoring data; through these technical means, the video surveillance system is made highly intelligent, automated and precise, thereby improving the monitoring efficiency, response speed and data utilization value of the system.

[0043] Furthermore, the algorithm collaboration platform also includes an analysis and detection module, which is used to perform machine learning based on the recognition results and the analysis and detection results to obtain an analysis and detection model; wherein, the recognition results are input into the analysis and detection model to output a reference analysis and detection result.

[0044] Through the algorithm collaborative platform of the integrated analysis and detection module, the startup module calls the preset algorithm to perform target recognition on the image sample data and obtain the recognition result, and the detection and analysis module judges the recognition result according to the preset analysis and judgment rules to obtain the analysis and judgment result, so as to realize in-depth analysis and decision support of video surveillance data; the analysis and detection module performs machine learning based on the recognition result and the analysis and judgment result to obtain the analysis and judgment detection model, and inputs the recognition result into the model to output the reference analysis and judgment result, so as to realize system self-learning and optimization; through these technical means, the video surveillance system can be realized with high intelligence, adaptability and accurate analysis and judgment ability, thereby improving the monitoring efficiency, decision-making quality and long-term operation stability of the system.

[0045] Furthermore, the scene point selection module is also used to save a third association table, which is used to save the third association between the preset algorithm and the algorithm type; the comprehensive video platform also includes an account storage module, and the algorithm collaboration platform also includes a scene feature module and an algorithm recommendation module; the account storage module is used to store the current account for logging into the comprehensive video platform, and the current account includes an account ID, enterprise type and enterprise information.

[0046] The scenario feature module is used to obtain the second scenario information of the current account on the algorithm collaboration platform, mark the second scenario information, and obtain a scenario label, which includes the industry and purpose of the scenario information; query the second association table according to the second scenario information, and determine the second algorithm type that has a second association relationship with the second scenario information.

[0047] The algorithm recommendation module includes an association storage submodule and a recommendation submodule. The association storage submodule is used to query the third association relationship table according to the second algorithm type to determine the second preset algorithm information having a third association relationship with the second algorithm type. The recommendation submodule is used to perform machine learning according to the scene label, the second preset algorithm information and the enterprise type to obtain an algorithm recommendation model; wherein, the enterprise type is input into the algorithm recommendation model to output a reference preset algorithm.

[0048] By constructing a comprehensive video platform and algorithm collaboration system including a scene point selection module, a point capture configuration module, an account storage module, a scene feature module and an algorithm recommendation module, the scene information of the current account is obtained and marked through the scene feature module to generate a scene label, and the algorithm recommendation model is obtained through the machine learning function of the algorithm recommendation module, so as to realize intelligent analysis of enterprise types and scene characteristics and personalized recommendation of algorithms; the third association relationship table is queried through the association storage submodule to determine the preset algorithm that matches the scene information, and the reference preset algorithm is output through the recommendation submodule to realize accurate matching and efficient deployment of the algorithm and the monitoring scene; through these technical means, the video surveillance system is realized with high intelligence, customization and automation, thereby improving the system's adaptability, monitoring efficiency and user satisfaction.

[0049] In some embodiments, the comprehensive video platform includes a background configuration module, and the background configuration module includes a rights-based and domain-based configuration submodule, wherein the rights-based and domain-based configuration submodule is used to configure permissions for the comprehensive video platform. Specifically, the rights-based and domain-based configuration submodule supports the division of account roles, and divides account roles with different permissions according to business needs. By setting the background configuration module on the comprehensive video platform, through the rights-based and domain-based configuration submodule, the refined management of permissions of the comprehensive video platform is achieved, the on-demand allocation and effective control of permissions are achieved, and the system security and the compliance of user operations are ensured.

[0050] In some embodiments, the integrated video platform includes a background configuration module, which includes a system configuration submodule, wherein the system configuration submodule is used to configure the system parameters and data interfaces of the integrated video platform. Specifically, the system configuration submodule supports viewing and modifying the system configuration, parameter configuration, dynamic configuration of the API interface, etc. in the system configuration. By setting the background configuration module on the integrated video platform and accurately adjusting the system configuration submodule, the flexible configuration of the system parameters and data interfaces of the integrated video platform can be achieved, thereby realizing personalized customization and optimization of the platform functions and improving the overall performance of the system and user experience.

[0051] Figure 2 FIG. 1 is a schematic diagram of the workflow of an algorithm collaboration system based on a comprehensive video platform provided in an embodiment of the present application, which is provided as an example and not as a limitation. Figure 2 As shown, in this embodiment, the workflow of the algorithm collaboration system based on the comprehensive video platform, the comprehensive video platform is deployed in the provincial comprehensive platform of the provincial branch, and the algorithm collaboration platform includes an algorithm verification module and an algorithm integration module, wherein the algorithm verification module is deployed in the provincial comprehensive platform of the provincial branch, and the algorithm integration module is deployed in the headquarters and the provincial branch respectively. And, a camera cluster is used to obtain video data, and, according to the acquisition task of the comprehensive video platform, the video data is intercepted to obtain the picture sample data, and the picture sample data includes picture data and / or real-time data stream. The camera cluster is deployed in the industry platform of the municipal branch.

[0052] The algorithm integration module includes an algorithm iteration module, which is used to update the preset algorithm; the algorithm iteration module is deployed at the headquarters. The algorithm integration module includes an algorithm service module (i.e., the aforementioned algorithm docking module), which is used to configure the preset algorithm according to the service specification so that the preset algorithm is adapted to the algorithm verification module; the algorithm service module is deployed at the provincial branch.

[0053] The specific workflow in this embodiment is roughly described as follows: (1) Headquarters: receive analysis results and materials; iterate algorithm; (2) Algorithm verification module: connect with the headquarters algorithm and display the results of the algorithm; call GPU resources for detection, and have the ability to issue capture tasks and allocate algorithms; have the function of judging true alarms and false alarms; automatically collect pictures; (3) Provincial comprehensive platform: connect with the algorithm verification module and receive capture tasks; issue capture tasks to the municipal platform; (4) Municipal comprehensive platform: gather the points of the municipal industry platforms; receive the capture tasks of the provincial comprehensive platform, and capture pictures according to the tasks; return the pictures to the provincial comprehensive platform.

[0054] The newly developed AI algorithm verification module is mainly used for the verification of the headquarters' self-developed algorithms. In addition to realizing specific functional modules such as algorithm configuration, alarm analysis, alarm display, and statistical analysis, two main docking tasks need to be realized at the same time. The first is to connect to a province's comprehensive video management platform, and by calling the comprehensive video management platform's capture service, realize the front-end camera's operation control, preset position setting and other capabilities, and obtain the front-end camera image data and real-time data stream. The second is to connect to the headquarters' AI algorithm, realize the call of the AI ​​algorithm for target recognition, and finally store the recognition results and camera screen sample data, manually judge the alarm results, and count the algorithm's operation quality and effect data, collect relevant materials, and form a conclusion report.

[0055] In this embodiment, by deploying the algorithm coordination system of the comprehensive video platform, the verification and iteration of the algorithm developed by the headquarters are realized, the adaptation and optimization problems of the algorithm in practical applications are solved, the accuracy and efficiency of the algorithm are improved, and the quality of intelligent analysis and judgment of video surveillance data is ensured. At the same time, through the connection of the algorithm verification module with the provincial comprehensive video management platform and the headquarters AI algorithm, the data collection of the front-end camera and the target recognition function of the AI ​​algorithm are realized, the real-time problem of data acquisition and processing is solved, and the automation level and response speed of the system are improved. In addition, through the statistical analysis module, the alarm results are manually judged and the statistical algorithm of the algorithm operation quality and effect data is collected, relevant materials are collected, and a conclusion report is formed to solve the problems of algorithm effect evaluation and continuous improvement, and improve the intelligence level and operation and maintenance management level of the entire video surveillance system.

[0056] Unify the algorithm verification process, establish a unified algorithm verification platform, realize functions such as algorithm configuration, alarm analysis, alarm display, and statistical analysis, and improve the efficiency and standardization of algorithm verification work. Convenient data acquisition, connect to the Hebei Provincial Comprehensive Video Management Platform, realize camera operation control, image data acquisition, and real-time data stream acquisition, and facilitate algorithm data collection and analysis. Efficient result analysis, provide manual analysis function, evaluate algorithm recognition results, and provide a basis for algorithm optimization. Data statistical analysis, statistics and analysis of algorithm operation quality, effect and other data, provide data support for algorithm iterative optimization. Algorithm iterative optimization, collect algorithm verification results and materials, provide data support for algorithm iterative optimization, and improve algorithm performance and robustness.

[0057] The overall implementation plan of the algorithm collaboration system based on the comprehensive video platform is divided into 8 steps:

[0058] 1. Algorithm service docking: docking and adapting with the headquarters algorithm service specifications and AI algorithm verification module to realize algorithm capture, detection and display;

[0059] 2. Scene point selection: Select points and scenes according to the algorithm category, and find scenes and points suitable for the algorithm to verify and iterate the algorithm;

[0060] 3. Point capture configuration: For the selected points, set the preset positions according to the adapted scenes, determine the capture rules and take the capture;

[0061] 4. AI algorithm verification module operation: After the algorithm service is connected, combined with the point capture rules, the AI ​​algorithm verification module is officially operated;

[0062] 5. Detection result analysis: The captured images are tested, and the results of the test are analyzed by dedicated algorithm analysis personnel according to the algorithm analysis rules;

[0063] 6. Test result statistics: Summarize and count the operation results of the algorithm within a unit time and form a report;

[0064] 7. Algorithm material collection: The operation results and materials are sent to the headquarters through designated channels, and the headquarters experts conduct special training and cooperate with the iteration and upgrade of the algorithm;

[0065] 8. Commercial promotion of algorithms: According to the requirements of the headquarters, provincial and municipal branches will simultaneously look for real customers and real scenarios for commercial use.

[0066] Through algorithm service docking, scene point selection and point capture configuration, this solution achieves the compatibility of the algorithm and the platform, the accuracy of scene matching and the standardization of image acquisition, and improves the efficiency of algorithm integration and verification. The operation of the AI ​​algorithm verification module and the judgment of the test results ensure the real-time monitoring and accurate evaluation of the algorithm performance, and improve the reliability of the system. The statistics of test results and the collection of algorithm materials provide data support for the continuous optimization of the algorithm. Finally, through the commercial promotion of the algorithm, the problems of market application landing are alleviated, the commercial value of the algorithm is enhanced, and the intelligence level of the overall monitoring system is improved.

[0067] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0068] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0069] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0070] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it can implement the steps in the above-mentioned embodiments of the method of the algorithm collaboration system based on the comprehensive video platform as mentioned above.

[0071] An embodiment of the present application provides a computer program product. When the computer program product runs on an electronic device, the electronic device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0072] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device capable of carrying the computer program code to an electronic device, a recording medium, a computer memory, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), an electric carrier signal, a telecommunication signal, and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0073] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0074] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0075] In the embodiments provided in the present application, it should be understood that the disclosed devices / network equipment and methods can be implemented in other ways. For example, the device / network equipment embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0076] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0077] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. An algorithm collaboration system based on a comprehensive video platform, characterized in that: include: An integrated video platform, used to obtain picture sample data from a camera cluster, and to provide a data interface for data interaction between the integrated video platform and the algorithm collaboration platform; The algorithm collaboration platform comprises an algorithm integration module and an algorithm verification module, wherein the algorithm integration module is used to provide a preset algorithm to the algorithm verification module; and the algorithm verification module is used to obtain the picture sample data from the data interface, call the preset algorithm to perform target recognition on the picture sample data, and obtain the analysis result of the picture sample data; the algorithm collaboration platform sends the analysis result corresponding to the picture sample data to the comprehensive video platform through the data interface; The camera cluster is used to acquire video data, and intercept the video data according to the acquisition task of the integrated video platform to obtain the picture sample data, wherein the picture sample data includes picture data and / or real-time data stream.

2. The algorithm collaboration system based on the comprehensive video platform as claimed in claim 1, characterized in that: The comprehensive video platform includes a storage module, and the storage module is used to store the picture sample data and the analysis result.

3. The algorithm collaboration system based on the comprehensive video platform as claimed in claim 2, characterized in that: The algorithm collaboration platform further includes an algorithm configuration module, which is used to configure the preset algorithm according to a preset scene; the camera cluster includes a first control module and more than four first camera devices; and the storage module further stores a first association relationship table, which is used to store a first association relationship between the preset algorithm information and the camera device information; After the preset algorithm configuration is completed, the algorithm collaboration platform sends configuration information to the integrated video platform, the configuration information including the first preset algorithm information; The integrated video platform queries the first association relationship table according to the configuration information to determine the first camera device information having the first association relationship with the first preset algorithm information; The integrated video platform sends the first acquisition task and the first camera device information to the camera cluster; The first control module is used to control the first camera device corresponding to the first camera device information, intercept the video data according to the first acquisition task, and obtain first picture sample data; and use the first picture sample data as the picture sample data.

4. The algorithm collaboration system based on the comprehensive video platform as claimed in claim 1, characterized in that: The algorithm integration module stores service specifications corresponding to the preset algorithm; the algorithm collaboration platform also includes an algorithm docking module, which is used to configure the preset algorithm according to the service specifications so that the preset algorithm is adapted to the algorithm verification module.

5. The algorithm collaboration system based on the comprehensive video platform as claimed in claim 4, characterized in that: The camera cluster includes a second control module and more than four second camera devices; the algorithm collaboration platform also includes a scene point selection module and a point capture configuration module, wherein: The scene point selection module is used to determine the first algorithm type according to the preset algorithm after the algorithm docking module configures the preset algorithm according to the service specification; the scene point selection module stores a second association relationship table, and the second association relationship table is used to save the second association relationship between the algorithm type and the point information and the scene information; The scene point selection module queries the second association relationship table according to the first algorithm type to determine the first point information and the first scene information having the second association relationship with the first algorithm type; The point capture configuration module is used to determine the preset position of the second camera device according to the first point information and the first scene information; The second control module is used to control the second camera device to move to the preset position, intercept the video data according to the second acquisition task, and obtain second picture sample data; and use the second picture sample data as the picture sample data.

6. The algorithm collaboration system based on the comprehensive video platform as claimed in claim 5, characterized in that: The algorithm collaboration platform also includes a startup module, a detection and analysis module, and a detection statistics module, wherein: The startup module is used to call the preset algorithm to perform target recognition on the picture sample data after the algorithm docking module configures the preset algorithm according to the service specification to obtain the recognition result of the picture sample data; The detection and judgment module is used to judge the recognition result according to the preset judgment rules to obtain the judgment result; The statistical display module is used to visualize the recognition result and the analysis result corresponding to the recognition result within a unit time.

7. The algorithm collaboration system based on the comprehensive video platform as claimed in claim 6, characterized in that: The algorithm collaboration platform also includes a research and judgment detection module, which is used to perform machine learning based on the recognition result and the research and judgment result to obtain a research and judgment detection model; Among them, the recognition result is input into the analysis and detection model, and a reference analysis and detection result can be output.

8. The algorithm collaboration system based on the comprehensive video platform as claimed in claim 5, characterized in that: The scene point selection module is also used to save a third association relationship table, and the third association relationship table is used to save a third association relationship between a preset algorithm and an algorithm type; the comprehensive video platform also includes an account storage module, and the algorithm collaboration platform also includes a scene feature module and an algorithm recommendation module; The account storage module is used to store the current account for logging into the comprehensive video platform, wherein the current account includes an account identifier, an enterprise type, and enterprise information; The scenario feature module is used to obtain the second scenario information of the current account on the algorithm collaboration platform, mark the second scenario information, and obtain a scenario tag, wherein the scenario tag includes the industry and purpose task of the scenario information; Querying the second association relationship table according to the second scene information to determine a second algorithm type having the second association relationship with the second scene information; The algorithm recommendation module includes an association storage submodule and a recommendation submodule, wherein the association storage submodule is used to query the third association relationship table according to the second algorithm type to determine the second preset algorithm information having the third association relationship with the second algorithm type; The recommendation submodule is used to perform machine learning based on the scenario label, the second preset algorithm information and the enterprise type to obtain an algorithm recommendation model; Among them, the enterprise type is input into the algorithm recommendation model, and a reference preset algorithm can be output.

9. The algorithm collaboration system based on the comprehensive video platform as claimed in claim 1, characterized in that: The comprehensive video platform includes a background configuration module, and the background configuration module includes a rights-based and domain-based configuration submodule, wherein the rights-based and domain-based configuration submodule is used to perform permission configuration on the comprehensive video platform.

10. The algorithm collaboration system based on the comprehensive video platform according to claim 1, characterized in that: The comprehensive video platform includes a background configuration module, and the background configuration module includes a system configuration submodule, wherein the system configuration submodule is used to configure system parameters and data interfaces of the comprehensive video platform.

Images