Video stream detection system, method and equipment in security cascade scene

Abstract

The invention relates to a video stream detection system, method and device in a security cascade scene. The system comprises an inspection center, a distributed detection cluster and to-be-detected video channel information, wherein the inspection center comprises at least two inspection nodes, and the distributed detection cluster comprises at least two detection nodes; the inspection node is used for issuing a detection task to the corresponding detection node according to the obtained customer demand; and the detection node is used for determining corresponding target to-be-detected video channel information from the to-be-detected video channel information based on the detection task, pulling and detecting a corresponding to-be-detected video based on the target to-be-detected video channel information, and pushing a detection result to the inspection center to generate a corresponding perception video detection result. The video detection method and device can adapt to various different application environments, the specification requirement of the server is lowered, and the video detection accuracy is improved.

Description

Technical Field

[0001] The present application relates to the field of video detection technology, and in particular to a video stream detection system, method and device in a security cascade scenario. Background Art

[0002] The video detection and analysis structure of the traditional multi-level video convergence monitoring platform includes: the district and county video access platform 1-N is responsible for placing the access and management capabilities of video equipment, and the internal streaming media service (or cluster) is responsible for forwarding the media data stream to multiple clients for pulling video streams for viewing or storage, which relieves the pressure of front-end equipment and saves the bandwidth of the network access layer. The municipal bureau video convergence platform then converges all video points in various districts and counties to shield various complex equipment interfaces for upper-level business applications. The intelligent operation and maintenance platform is located at the upper level and uses all video points of the municipal bureau convergence platform for operation and maintenance services, such as video detection and analysis services. However, if video stream / recording stream detection and analysis is performed on the municipal bureau platform, due to problems such as long links, unstable networks, and large network resource usage, the municipal bureau platform will be overloaded and unable to view real-time videos. At the same time, the detection results may also be inaccurate, affecting the normal use of upper-level business functions.

[0003] At present, no effective solution has been proposed for the problems of high requirements on server resources and poor adaptability to different environments in the existing technology. Summary of the invention

[0004] Based on this, it is necessary to provide a video stream detection system, method and device in a security cascade scenario to address the above technical problems.

[0005] In a first aspect, the present application provides a video stream detection system in a security cascade scenario. The system includes: an inspection center, a distributed detection cluster, and video channel information to be detected; wherein the inspection center includes at least two inspection nodes, and the distributed detection cluster includes at least two detection nodes; Inspection nodes are used to send inspection tasks to corresponding inspection nodes based on acquired customer needs; The detection node is used to determine the corresponding target video channel information to be detected from the video channel information to be detected based on the detection task, pull the corresponding video to be detected based on the target video channel information to be detected and perform detection, and push the detection result to the inspection center to generate the corresponding perception video detection result.

[0006] In one embodiment, the system further comprises a dispatch center, wherein the dispatch center communicates with the inspection center; The dispatch center is used to set up scheduled video stream detection tasks; The dispatch center is also used to instruct the corresponding inspection node to start executing the video stream detection task when the scheduled video stream detection task is triggered.

[0007] In one embodiment, the system further includes a data transfer cluster and a data storage cluster; wherein the inspection center is connected to the data transfer cluster and the data storage cluster respectively; Data storage cluster, used to persistently store the detection rule information for each detection task; Data transfer cluster, used to transmit detection results.

[0008] In one of the embodiments, the inspection node is further used to automatically send a re-inspection task to a corresponding inspection node based on the inspection result after generating the inspection result; The re-inspection task includes determining abnormal video results based on the detection results, and re-inspecting the abnormal video results in a loop until the number of re-inspections reaches a preset re-inspection threshold to obtain the final detection result, wherein as the number of cycles increases, the number of video streams re-inspected in each round decreases.

[0009] In one of the embodiments, the inspection center is also used to obtain the platform domain information and IP information of each detection node, and match the video channel information to be detected with the detection node based on the platform domain information and IP information, determine the target detection node that matches the video channel information to be detected from all detection nodes, and issue the detection task for the corresponding video channel information to be detected to the target detection node.

[0010] In one of the embodiments, the inspection center is further used to traverse each detection node to determine whether there is an abnormal state in each detection node; The inspection center is also used to match the to-be-detected video channel information corresponding to the abnormal detection node with the nodes other than the abnormal detection node in the detection nodes to determine the final detection node when an abnormal detection node in an abnormal state is detected among the detection nodes; The inspection center is also used to send the corresponding video channel information to be inspected to the final inspection node.

[0011] In one of the embodiments, the distributed detection cluster is also used to traverse each detection node, and after detecting that all detection nodes have completed the detection of the video to be detected corresponding to the video channel information to be detected, the detection task is ended and the detection result is returned to the inspection center.

[0012] In a second aspect, the present application also provides a video stream detection method in a security cascade scenario. The method is applied to the above system, including: Through at least two inspection nodes included in the inspection center, detection tasks are issued to the detection nodes in the corresponding distributed detection cluster according to the obtained customer needs; wherein the detection node determines the corresponding target video channel information to be detected from the video channel information to be detected, pulls the corresponding video to be detected based on the target video channel information to be detected and performs detection, pushes the detection result to the inspection center, and generates the corresponding perception video detection result.

[0013] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program: Through at least two inspection nodes included in the inspection center, detection tasks are issued to the detection nodes in the corresponding distributed detection cluster according to the obtained customer needs; wherein the detection node determines the corresponding target video channel information to be detected from the video channel information to be detected, pulls the corresponding video to be detected based on the target video channel information to be detected and performs detection, pushes the detection result to the inspection center, and generates the corresponding perception video detection result.

[0014] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented: Through at least two inspection nodes included in the inspection center, detection tasks are issued to the detection nodes in the corresponding distributed detection cluster according to the obtained customer needs; wherein the detection node determines the corresponding target video channel information to be detected from the video channel information to be detected, pulls the corresponding video to be detected based on the target video channel information to be detected and performs detection, pushes the detection result to the inspection center, and generates the corresponding perception video detection result.

[0015] The above-mentioned video stream detection system, method and equipment in a security cascade scenario, the system includes, an inspection center, a distributed detection cluster and video channel information to be detected, the inspection center includes multiple inspection nodes, and the distributed detection cluster includes multiple detection nodes; the inspection node is used to issue detection tasks to the corresponding detection nodes according to customer needs; the detection node is used to determine the corresponding target video channel information to be detected from the video channel information to be detected according to the detection task, pull the corresponding video to be detected based on the target video channel information to be detected, and push the detection result to the inspection center to generate the corresponding perception video detection result. The structure of the video stream detection system is improved through this application, so that it can adapt to a variety of different application environments, and the specification requirements for the detection and analysis node server in this application are low, reducing the investment in hardware resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the structure of a video stream detection system in one embodiment; Figure 2 A schematic diagram of the structure of a video stream detection system in another embodiment; Figure 3 A schematic diagram of a retest result in an embodiment; Figure 4 A schematic diagram of a video stream detection method flow in a preferred embodiment; Figure 5 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0018] In one embodiment, Figure 1 As shown, a video stream detection system in a security cascade scenario is provided, including an inspection center, a distributed detection cluster, and video channel information to be detected; wherein the inspection center includes at least two inspection nodes 11, and the distributed detection cluster includes at least two detection nodes 12; Inspection node 11 is used to send inspection tasks to corresponding inspection nodes 12 according to acquired customer needs; The detection node 12 is used to determine the corresponding target video channel information to be detected from the video channel information to be detected based on the detection task, pull the corresponding video to be detected based on the target video channel information to be detected and perform detection, and push the detection result to the inspection center to generate the corresponding perception video detection result.

[0019] Specifically, in order to solve the performance problem of video stream detection and analysis collected by a large number of video devices in this application, the inspection center supports multi-node deployment to improve the reliability and throughput of the detection and analysis result processing, and the video detection module also adopts a distributed detection cluster to improve the concurrent performance and reliability of video streaming and detection analysis. The inspection center is connected to the client, obtains the customer demand issued by the client, and issues the detection task to the corresponding detection node 12 according to the customer demand. In actual applications, the customer demand is mostly a video detection and analysis instruction issued by the client, wherein the detection node 12 is mostly a different device for detecting the video stream belonging to the distributed detection cluster, and the configuration information of the video detection and analysis required (configuration information such as detection indicators, detection equipment information, detection trigger time, etc.) is mostly configured by relevant technical personnel. In actual applications, the above-mentioned detection node 12 is generally deployed as a horizontal multi-node to form the above-mentioned distributed detection cluster, that is, multiple servers are concentrated together, and each server implements the same business. The main role of the existence is to disperse the execution of tasks to improve the detection performance. At the same time, if a detection node fails, it can also support the node to fail over, thereby ensuring the reliability of video detection and analysis. The above detection tasks are mostly issued by users, and a batch or a certain type of collected video stream is detected. Further, in this embodiment, the detection task can be sent to a suitable detection node 12, such as the detection task of a certain area can be sent to the detection node 12 closest to the area, or the detection task of a certain area can be sent to the detection node 12 with the best performance, etc., and the allocation of the detection task can be determined by relevant technical personnel according to the actual situation in practical application. It should be noted that the present application is not to directly allocate the video to be detected, but to allocate the video channel information to be detected to the detection node 12, wherein the video channel information to be detected includes but is not limited to the video acquisition device encoding, video device access domain, video device label, video device IP address, etc. After the detection node 12 obtains the target video channel information to be detected corresponding to the detection task, the detection node 12 pulls the corresponding video resource to be detected from the streaming media cluster according to the above target video channel information to be detected, and detects it. Finally, the detection result is pushed to the inspection center, and the inspection center generates the corresponding perception video detection result, which can be a standard visualization report generated by the inspection center, etc., and is fed back to the client for users to view. In practical applications, it is generally necessary to detect a large amount of video channel information to be detected. At this time, different video channel information to be detected can be allocated to different detection nodes 12 according to actual needs, thereby improving detection efficiency and accuracy.

[0020] Through this application, by deploying a multi-node inspection center and a multi-node detection cluster, it is possible to avoid gathering all the videos to be detected at one point / platform, which can effectively avoid the problem of excessive pressure on the platform. In addition, in this application, the channel information of the video to be detected is allocated instead of directly allocating the video to be detected. On the basis of improving the overall detection efficiency of the video, the hardware requirements for each node are reduced, and the detection efficiency is also improved, thereby avoiding affecting the normal use of the upper-level business functions.

[0021] In one embodiment, the system further includes a dispatch center 21, wherein the dispatch center 21 communicates with the inspection center; The dispatch center 21 is used to set the timed video stream detection task; The dispatch center 21 is also used to instruct the corresponding inspection node to start executing the video stream detection task when the scheduled video stream detection task is triggered.

[0022] Specifically, the above inspection nodes are all used to be deployed on the server in a microservice manner, and preferably, there is no functional difference between the inspection nodes. The dispatch center 21 and the inspection center can communicate with each other. In actual applications, the inspection center can register the dispatch task with the dispatch center 21 according to the customer needs obtained from the client, and the dispatch center 21 is responsible for setting the detection task. The detection task can be set at a time, such as being set to trigger the corresponding task immediately when receiving the video stream detection task, or setting the trigger time according to actual needs, such as being configured to execute the video stream detection task at 8:55 every morning, or being configured to execute at 2:10 next week, etc. When the scheduled video stream detection task set by the dispatch center 21 is triggered, it is fed back to the corresponding inspection node, indicating that the inspection node starts to execute the corresponding video stream detection task. Through the cooperation of the dispatch center 21 and the inspection center, flexible detection of large quantities of videos can be achieved, so that the system can adapt to different application scenarios more flexibly.

[0023] In one embodiment, the system further includes a data transfer cluster and a data storage cluster; wherein the inspection center is connected to the data transfer cluster and the data storage cluster respectively; Data storage cluster, used to persistently store the detection rule information for each detection task; Data transfer cluster, used to transmit detection results.

[0024] Specifically, Figure 2The diagram is a schematic diagram of the structure of a video stream detection system in an embodiment, and the inspection center is connected to the data storage cluster and the data transfer cluster respectively. The data storage cluster usually also includes multiple data storage nodes 23, and different data storage nodes 23 are used to persistently store the configuration information of the task to be detected, the final video detection results and other data, and the embodiment includes multiple data storage nodes 23. If a faulty node occurs, other nodes can take over the work, thereby forming a highly reliable data storage capability to avoid losses caused by single point failures. Among them, the detection rule information of the detection task includes but is not limited to determining the resources to be detected that need to be detected, clarifying the detection index items, whether re-inspection is required, re-inspection strategy, trigger time, etc. In some embodiments, persistent storage of data can also be achieved based on the data storage cluster. The above-mentioned data transfer cluster usually also includes multiple data transfer nodes 22. Each data transfer node 22 is used for interactive decoupling of data between the inspection node and the detection node to reduce inter-service dependency. In practical applications, the capabilities of each data transfer node 22 are basically the same. Setting multiple data transfer nodes 22 can improve the data transmission capability, and the data transfer cluster can also assist in the transmission of asynchronous messages. Asynchronous message transmission can achieve message decoupling between services, improve detection performance and reduce inter-service dependency. After completing the analysis of the video to be detected, the detection results are pushed to the intermediate storage medium, that is, the above-mentioned data transfer cluster. Then the inspection service is responsible for persistently storing the detection results, playing a role in the message transmission collaboration capability between multiple services during the video stream detection and analysis process. Among them, pushing data to the inspection center through the data transfer cluster can reduce the pressure on the inspection node to receive the detection results generated by the detection node, and can also achieve asynchronous processing capabilities. It is understandable that if the detection node pushes the detection results directly to the inspection node, the inspection node must ensure that no abnormalities occur, which greatly increases the requirements for the inspection node. If the detection results are pushed directly to the inspection node, the inspection node needs to perform additional logical processing, which will also slow down the performance of the detection node.

[0025] In one of the embodiments, the inspection node is further used to automatically send a re-inspection task to a corresponding inspection node based on the inspection result after generating the inspection result; The re-inspection task includes determining abnormal video results based on the detection results, and re-inspecting the abnormal video results in a loop until the number of re-inspections reaches a preset re-inspection threshold to obtain the final detection result, wherein as the number of cycles increases, the number of video streams re-inspected in each round decreases.

[0026] Specifically, in actual applications, there will be some interference factors, such as network jitter, high streaming media pressure and other problems, which may cause misdetection with a high probability. In order to improve the detection accuracy, the inspection node in this embodiment also supports the automatic re-inspection function. After each round of detection task is completed, the inspection service can re-issue the detection according to the detection results and pre-configured resources, and select the abnormal video results that need to be re-inspected from the detected abnormal results for cyclic re-inspection until the number of re-inspections reaches the re-inspection number threshold, and obtain the final detection result, wherein the screening method for abnormal video results includes but is not limited to, after obtaining the detection result through the first video detection task, randomly extracting a preset number of video stream data from the abnormal detection result as an abnormal video result for re-inspection; or extracting abnormal video results from the detection result according to a preset screening strategy, the screening strategy includes but is not limited to, detecting the environment of the video in the detection result, if the environmental factor meets the preset environmental requirements, it is determined as an abnormal video result and re-inspected, etc. It can be understood that the screening method of abnormal video results is not limited in this embodiment, and all methods that can determine abnormal video results from the detection results should fall within the scope of protection of this application.

[0027] Furthermore, the re-inspection task in this embodiment usually includes multiple rounds of re-inspection tasks, and as the number of cycles increases, the number of video streams re-inspected in each round decreases gradually, that is, after completing the first round of detection of the video to be detected, the videos that need to be re-inspected for the second round are screened out from the detection results of the first round, and after obtaining the second round of detection results for the second round of re-inspected videos, the videos that need to be re-inspected for the third round are screened out from the second round of detection results… Repeat the above-mentioned cyclic re-inspection method until the number of re-inspections reaches a preset re-inspection number threshold (such as 3 times), and it should be noted that, considering the need to take into account both detection accuracy and detection speed in actual applications, the number of re-inspected videos in each round should be less than the number of re-inspected videos in the previous round. For example, if 150 abnormal detection results are obtained through the first video detection task, the second round can be set to randomly extract 100 video stream data from these 150 abnormal detection results for re-inspection. If 50 abnormal results are obtained through the second round of re-inspection tasks, the third round can be set to extract 20 of these 50 abnormal results for re-inspection... until the number of re-inspections reaches the preset re-inspection threshold, where the re-inspection threshold can be set to 3 times. In some preferred embodiments, when performing re-inspection tasks, it also supports the configuration of re-inspection trigger time (such as starting re-inspection at 13:00, or re-inspection 2 hours after obtaining the detection results, etc.), and also supports the configuration of re-inspection resources, etc. The automatic re-inspection task can significantly improve the accuracy of video detection analysis and reduce the impact of interference factors on detection results. Figure 3In one embodiment, after two rounds of re-inspection, the accuracy rate is nearly doubled (the green arrow in the figure is the newly added normal number of re-inspection). In practical applications, the re-inspection object can be flexibly selected according to actual needs, such as the channel that needs to be re-inspected. For example, the full channel is to re-inspect the videos to be detected corresponding to all channels. For example, the abnormal channel is to re-inspect only the videos to be detected that show abnormalities after detection. Similarly, the normal channel is to re-inspect the videos to be detected that show normal after detection. Further, the number of re-inspections can be set by relevant technical personnel according to actual needs, such as 3 times; further, the start time of the re-inspection can also be set by oneself, such as it can be set to start the re-inspection immediately after the task (such as the video detection task in this article) is completed, or it can be set to start the re-inspection at a specified time, etc.

[0028] In one of the embodiments, the inspection center is also used to obtain the platform domain information and IP information of each detection node, and match the video channel information to be detected with the detection node based on the platform domain information and IP information, determine the target detection node that matches the video channel information to be detected from all detection nodes, and issue the detection task for the corresponding video channel information to be detected to the target detection node.

[0029] In the prior art, for scenarios where multiple video access platforms are cascaded and video stream detection and analysis is performed at the upper level (such as the city-district-county network deployment architecture in the prior art), simply increasing the number of concurrent video stream detection and analysis will have a significant impact on the performance of the video access platform, affecting video stream preview and network bandwidth resource usage, and ultimately leading to increased on-site costs for customers.

[0030] In some preferred embodiments, the present application can calculate the concurrency of each node according to the specification information of each node, wherein the specification information includes but is not limited to the memory, CPU / GPU parameters, operating system, remaining detection and analysis quantity of the node, and the specification information contains the basic parameter information of the corresponding detection node. After matching the video channel information to be detected with the concurrency and the detection node, a target detection node that matches the video channel information to be detected is determined from all detection nodes, and a detection task for the corresponding video channel information to be detected is issued to the target detection node, wherein the matching process includes matching the platform domain information and IP network segment information of the detection node and the video stream to be detected. Specifically, this embodiment follows the principle of nearest matching, and based on the platform domain information and IP network segment information of the detection node and the video channel information to be detected, the optimal path allocation and issuance with respect to the video channel information to be detected is screened out based on the detection node and the video channel information to be detected. Specifically, matching based on platform domain information includes grouping detection nodes by platform domain, and grouping the video channel information to be detected by platform domain, and then matching detection nodes in the same domain with the video channel information to be detected; similarly, matching based on IP information includes grouping detection nodes by IP segments, and grouping the video channel information to be detected by platform IP segments, and then matching detection nodes in the same domain with the video channel information to be detected. This embodiment can realize the principle of nearest matching of the video channel information to be detected, and save network traffic bandwidth and ensure the stability of the video access platform environment while meeting the requirements of video stream (real-time stream / recorded stream) detection and analysis in large scenarios. In addition, this application has low requirements for the specifications of the detection node server, and dynamically adapts the detection capability through the video detection component to identify the server specifications, thereby reducing the investment in hardware resources.

[0031] In one of the embodiments, the inspection center is further used to traverse each detection node to determine whether there is an abnormal state in each detection node; The inspection center is also used to match the to-be-detected video channel information corresponding to the abnormal detection node with the nodes other than the abnormal detection node in the detection nodes to determine the final detection node when an abnormal detection node in an abnormal state is detected among the detection nodes; The inspection center is also used to send the corresponding video channel information to be inspected to the final inspection node.

[0032] Specifically, the inspection center traverses each detection node to detect whether there is an abnormal detection node in an abnormal state in each detection node, wherein detection node abnormalities will occur when a server abnormality, server restart, network disconnection, power outage, etc. occurs. If an abnormal detection node is detected, the video channel information to be detected corresponding to the abnormal detection node is matched with other normal detection nodes, wherein the matching method is as shown in the detection node and the video channel information to be detected above, and the video channel information to be detected is sent down according to the matched optimal path, and the video channel information to be detected on the abnormal detection node is allocated to other normal nodes for continued detection. This embodiment can improve the ability of fault migration and self-healing after an abnormality occurs in the video resource detection process, and ensure the high reliability execution of the video detection task.

[0033] In one of the embodiments, the distributed detection cluster is also used to traverse each detection node, and after detecting that all detection nodes have completed the detection of the video to be detected corresponding to the video channel information to be detected, the detection task is ended and the detection result is returned to the inspection center.

[0034] Specifically, the distributed detection cluster traverses each detection node, cleans up the subtasks of each detection node, and after detecting that all detection nodes have completed the detection of the video to be detected, exits the detection task and returns the detection result to the inspection center.

[0035] Based on the same inventive concept, the embodiment of the present application also provides a video stream detection method for implementing the video stream detection system involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more video stream detection method embodiments provided below can refer to the limitations on the video stream detection system above, and will not be repeated here.

[0036] In one embodiment, a video stream detection method is provided, which is applied in the video stream detection system as described above, and the method includes: Through at least two inspection nodes included in the inspection center, detection tasks are issued to the detection nodes in the corresponding distributed detection cluster according to the obtained customer needs; wherein the detection node determines the corresponding target video channel information to be detected from the video channel information to be detected, pulls the corresponding video to be detected based on the target video channel information to be detected and performs detection, pushes the detection result to the inspection center, and generates the corresponding perception video detection result.

[0037] The present application also provides a preferred embodiment of a video stream detection method in a security cascade scenario. Figure 4 The figure is a flow chart of a video stream detection method in a preferred embodiment.

[0038] S510, the distributed detection cluster calculates the concurrency of the video to be detected according to the detection node server specifications (such as memory, CPU / GPU, operating system type, remaining detection analysis number, etc.).

[0039] S520, determine the relationship between the detection node and the video channel information to be detected. The calculation of the matching relationship can be determined from two aspects: if the detection node and the video channel information to be detected have been configured with platform domain information, the detection nodes are grouped by platform domain, and the video channel information to be detected is grouped by platform domain, and then the detection nodes and video channel information to be detected under the same domain are matched; if the detection node and the video channel information to be detected are not configured with platform domain information, the detection nodes are grouped by IP segments, and the video channel information to be detected is grouped by platform IP segments, and then the detection nodes and video channel information to be detected under the same IP are matched.

[0040] S530: Allocate the to-be-detected video channel information to corresponding detection nodes according to the concurrency number and the relationship degree.

[0041] S540, asynchronously triggering all detection nodes.

[0042] S550, detect whether each detection node has completed the detection task normally. If an abnormal detection node is detected, the video channel information to be detected corresponding to the abnormal detection node is re-matched with other normal detection nodes, wherein the matching method is as shown in the detection node and the video channel information to be detected above, so that the detection of the video to be detected is completed through other normal detection nodes. If it is detected that all detection nodes have completed the detection task normally, the detection node exits and returns the detection result to the inspection center.

[0043] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0044] Each module in the above video stream detection system can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0045] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data related to the video stream detection algorithm. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a video stream detection method is implemented.

[0046] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0047] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: Through at least two inspection nodes included in the inspection center, detection tasks are issued to the detection nodes in the corresponding distributed detection cluster according to the acquired customer needs; wherein, the detection node determines the corresponding target video channel information to be detected from the video channel information to be detected, pulls the corresponding video to be detected based on the target video channel information to be detected and performs detection, pushes the detection result to the inspection center, and generates the corresponding perception video detection result.

[0048] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0049] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0050] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0051] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A video stream detection system in a security cascade scenario, characterized in that: The system includes: an inspection center, a distributed detection cluster, and video channel information to be detected; wherein the inspection center includes at least two inspection nodes, and the distributed detection cluster includes at least two detection nodes; The inspection node is used to send inspection tasks to the corresponding inspection node according to the acquired customer needs; The detection node is used to determine the corresponding target video channel information to be detected from the video channel information to be detected based on the detection task, pull the corresponding video to be detected based on the target video channel information to be detected and perform detection, and push the detection result to the inspection center to generate the corresponding perception video detection result.

2. The video stream detection system according to claim 1, characterized in that: The system further comprises a dispatching center, wherein the dispatching center communicates with the inspection center; The dispatch center is used to set a timed video stream detection task; The dispatch center is further used to instruct the corresponding inspection node to start executing the video stream detection task when the scheduled video stream detection task is triggered.

3. The video stream detection system according to claim 1, characterized in that: The system also includes a data transfer cluster and a data storage cluster; wherein the inspection center is connected to the data transfer cluster and the data storage cluster respectively; The data storage cluster is used to persistently store the detection rule information for each of the detection tasks; The data transfer cluster is used to transmit the detection result.

4. The video stream detection system according to claim 1, characterized in that: The inspection node is also used to, after generating the detection result, automatically issue a re-inspection task to the corresponding inspection node based on the detection result; Among them, the re-inspection task includes determining the abnormal video results based on the detection results, and performing cyclic re-inspections on the abnormal video results until the number of re-inspections reaches a preset re-inspection threshold to obtain the final detection results, wherein as the number of cycle rounds increases, the number of video streams re-inspected in each round decreases gradually.

5. The video stream detection system according to claim 1, characterized in that: The inspection center is also used to obtain the platform domain information and IP information of each of the detection nodes, and match the video channel information to be detected with the detection node based on the platform domain information and the IP information, determine the target detection node that matches the video channel information to be detected from all the detection nodes, and issue the detection task for the corresponding video channel information to be detected to the target detection node.

6. The video stream detection system according to claim 5, characterized in that: The inspection center is further used to traverse each of the detection nodes to determine whether there is an abnormal state in each detection node; The inspection center is further configured to, when detecting that there is an abnormal detection node in an abnormal state among the detection nodes, match the to-be-detected video channel information corresponding to the abnormal detection node with the nodes among the detection nodes other than the abnormal detection node to determine a final detection node; The inspection center is also used to send the corresponding video channel information to be detected to the final detection node.

7. The video stream detection system according to any one of claims 1 to 6, characterized in that: The distributed detection cluster is also used to traverse each of the detection nodes, and after detecting that all of the detection nodes have completed the detection of the video to be detected corresponding to the video channel information to be detected, end the detection task and return the detection result to the inspection center.

8. A video stream detection method in a security cascade scenario, characterized in that: The method is applied to the system according to any one of claims 1 to 7, and the method comprises: Through at least two inspection nodes included in the inspection center, detection tasks are issued to the detection nodes in the corresponding distributed detection cluster according to the acquired customer needs; wherein, the detection node determines the corresponding target video channel information to be detected from the video channel information to be detected, pulls the corresponding video to be detected based on the target video channel information to be detected and performs detection, pushes the detection result to the inspection center, and generates the corresponding perception video detection result.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the functions of the video stream detection system according to any one of claims 1 to 7 are realized.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the functions of the video stream detection system according to any one of claims 1 to 7 are realized.

Images