Video data analysis method, related equipment and storage medium

Abstract

The invention discloses a video data analysis method, related equipment and a storage medium, and the method comprises the steps: carrying out the basic analysis based on video data, and obtaining a basic analysis result; in response to the basic analysis result, representing that a target event is detected, calling an analysis plan associated with the target event, and performing advanced analysis on the video data based on the analysis plan to obtain an advanced analysis result; wherein the analysis plan comprises at least one analysis task representing advanced analysis, and the computing power resource required by the basic analysis is less than the computing power resource required by the advanced analysis. The advanced analysis is carried out only when the target event is detected, and more computing power resources are not occupied all the time when the target event is not detected, so that the computing power resources are saved; when the target event is detected, the associated analysis plan can be called to carry out advanced analysis to obtain the advanced analysis result, a further analysis result can be obtained through advanced analysis when needed, and event missing report and false report can be reduced.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a video data analysis method and related equipment and storage media. Background Art

[0002] Currently in the field of video data processing, by analyzing the video, it is possible to report when a corresponding event occurs.

[0003] However, different analysis levels require different computing resources. If you keep enabling analysis functions at all levels for a video, it may lead to a waste of computing resources. However, if you only enable basic analysis functions for a video, you will get too little intelligent analysis information, resulting in missed events and false alarms.

[0004] Therefore, how to reduce the waste of computing resources in the video analysis process while reducing missed events and false alarms has become an urgent problem to be solved. Summary of the invention

[0005] The present application at least provides a video data analysis method and related equipment and storage medium to reduce the waste of computing resources during the video analysis process while reducing event omissions and false alarms.

[0006] A first aspect of the present application provides a video data analysis method, comprising performing basic analysis based on video data to obtain basic analysis results; in response to the basic analysis results indicating that a target event has been detected, calling an analysis plan associated with the target event, and performing advanced analysis on the video data based on the analysis plan to obtain an advanced analysis result; wherein the analysis plan includes at least one analysis task representing the advanced analysis, and the computing power resources required for the basic analysis are less than the computing power resources required for the advanced analysis.

[0007] Therefore, first, basic analysis is performed based on the video data to obtain basic analysis results. When the basic analysis results indicate that a target event has been detected, the associated analysis plan is called for advanced analysis. Since advanced analysis is performed only when a target event is detected, it will not occupy more computing resources when no target event is detected (that is, when advanced analysis is not needed), thereby saving computing resources. At the same time, when a target event is detected, the associated analysis plan can be called for advanced analysis to obtain advanced analysis results. Further analysis results can be obtained through advanced analysis when needed, which can reduce missed reports and false alarms.

[0008] A second aspect of the present application provides an electronic device, comprising a memory and a processor coupled to each other, wherein the processor is used to execute program instructions stored in the memory to implement the video data analysis method in the first aspect.

[0009] A third aspect of the present application provides a computer-readable storage medium having program instructions stored thereon, and when the program instructions are executed by a processor, the video data analysis method in the first aspect described above is implemented.

[0010] The above scheme first performs basic analysis based on video data to obtain basic analysis results. When the basic analysis results indicate that a target event has been detected, the associated analysis plan is called to perform advanced analysis. Since advanced analysis is performed only when a target event is detected, it will not occupy more computing resources when no target event is detected (i.e., when advanced analysis is not needed), thereby saving computing resources. At the same time, when a target event is detected, the associated analysis plan can be called to perform advanced analysis to obtain advanced analysis results. Further analysis results can be obtained through advanced analysis when needed, which can reduce missed reports and false alarms.

[0011] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and are used together with the specification to illustrate the technical solution of the present application.

[0013] Figure 1 It is a flow chart of an embodiment of a video data analysis method of the present application; Figure 2 It is a flowchart of another embodiment of the video data analysis method of the present application; Figure 3 It is a schematic diagram of the framework of an embodiment of the scheduling management system of the present application; Figure 4 It is a schematic diagram of the framework of an embodiment of a video data analysis device of the present application; Figure 5 It is a schematic diagram of the framework of an embodiment of the electronic device of the present application; Figure 6 A schematic diagram of a framework of an embodiment of a computer-readable storage medium of the present application. DETAILED DESCRIPTION

[0014] The scheme of the embodiment of the present application is described in detail below in conjunction with the drawings of the specification.

[0015] In the following description, for the purpose of explanation rather than limitation, specific details such as specific system structures, interfaces, and technologies are provided to facilitate a thorough understanding of the present application.

[0016] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the objects associated before and after are in an "or" relationship. In addition, "many" in this article means two or more than two. In addition, the term "at least one" in this article means any combination of at least two of any one or more of a plurality of, for example, including at least one of A, B, and C, can mean including any one or more elements selected from the set consisting of A, B, and C.

[0017] Analyzing a video can obtain information of interest from the video, such as whether a target object appears in the picture. Video data can usually be obtained from multiple video channels, and multiple video channels can provide different video data streams. For example, multiple video channels provide video data streams shot by different cameras. When a video channel turns on the analysis function, the video data stream of the video channel will be sent to the corresponding analyzer. When the analyzer detects the corresponding target event, it will report the target event and generate the corresponding smart frame superimposed on the video data stream. After receiving the target event, the upper layer will perform corresponding processing.

[0018] Usually, each analysis function can only analyze the video data stream independently, and needs to be bound to the video channel when it is turned on. Therefore, after the analysis function is turned on, each analyzer can only analyze the video data stream of the video channel bound to it. At the same time, the video analysis capabilities of the device usually have performance limits, and usually after the analysis function is turned on, it will always analyze the video data stream of a single video channel, regardless of whether the video data stream is worth analyzing, which will cause a waste of computing resources. However, if only basic analysis is performed on the video data stream of the video channel in order to save computing resources, sufficient analysis results cannot be obtained, resulting in missed events and false alarms.

[0019] To this end, this application provides a video data analysis method, please refer to Figure 1 , the video data analysis methods include: Step S20: Perform basic analysis based on the video data to obtain basic analysis results.

[0020] Step S40: In response to the basic analysis result indicating that a target event is detected, an analysis plan associated with the target event is called, and an advanced analysis is performed on the video data based on the analysis plan to obtain an advanced analysis result.

[0021] The analysis plan includes at least one analysis task representing an advanced analysis, and the computing power resources required for the basic analysis are less than the computing power resources required for the advanced analysis.

[0022] The video data analysis method can be applied to camera equipment, which refers to equipment with camera functions and data processing functions, including but not limited to cameras deployed on urban roads, smart phones and other equipment. The video data analysis method can also be applied to servers, which refers to equipment with data processing functions, such as computers.

[0023] The analysis function can be preliminarily divided into basic analysis and advanced analysis according to the degree of demand for computing resources. Specifically, the computing resources required for basic analysis are less than those required for advanced analysis. Basic analysis can be an analysis function that requires lower computing resources, which can cover all video channels and perform basic analysis on the video data streams provided by all video channels without affecting the operation of the equipment. For example, the basic analysis function can include line intrusion, area intrusion, intelligent dynamic detection, etc. Advanced analysis can be further divided into channel-level analysis functions with higher demand for computing resources and device-level analysis functions with extremely high demand for computing resources. For example, the channel-level analysis function can be enabled for some video channels at the same time, such as target appearance recognition, target shape detection, etc.; while the device-level analysis function can only be enabled for a single video channel, that is, after a certain video channel enables the device-level analysis function, other video channels can no longer enable the device-level analysis function, or it is an analysis function that is only suitable for running when the device is idle, such as video structuring, all things detection, etc.

[0024] It can also be preliminarily divided into basic analysis and advanced analysis according to the characteristics of the analysis function itself. For example, basic analysis can be based on simple rule lines to identify dynamic behaviors associated with rule lines appearing in the video, such as line intrusion, area intrusion, intelligent motion detection, etc. Advanced analysis can further include precise analysis and full analysis. Precise analysis usually has an intelligent model for an object or behavior trained by a deep learning algorithm. During the analysis process, algorithms and models are used to accurately identify the target object or behavior, such as target appearance recognition and target shape detection. Full analysis usually has a wider recognition range than precise analysis, and can integrate various information to identify various objects and other behavior information in the video, thereby providing more effective information.

[0025] As mentioned above, basic analysis may include tripwire intrusion, area intrusion, intelligent dynamic detection, etc. Taking tripwire intrusion as an example, tripwire intrusion refers to the detection of the behavior of the target object crossing the specified boundary line. Therefore, when the basic analysis includes tripwire intrusion, if the basic analysis result indicates that the target event is detected, it means that the target object is detected crossing the specified boundary line, that is, the target event is the target object crossing the specified boundary line.

[0026] After a target event is detected, an analysis plan associated with the target event is called. Different target events can be associated with different analysis plans, or the same target event can be associated with multiple analysis plans.

[0027] Exemplarily, the analysis plan includes a target appearance analysis plan, which includes a target appearance detection task and a target appearance feature extraction task that represent advanced analysis. That is to say, after the basic analysis result represents the detection of a target event, the target appearance analysis plan can be called to perform target appearance detection and target appearance feature extraction on the video data; after obtaining the target appearance features, the target appearance features can be compared with a preset feature library, and if they are no longer in the preset feature library, they are marked as suspicious targets. As another example, the analysis plan can also include a target appearance recognition plan, which includes a target appearance detection task, a target appearance feature extraction task, and a target appearance comparison task. As yet another example, the analysis plan can also include a target shape detection task and a target appearance feature extraction task. As another example, the analysis plan may also include global target recognition, which specifically includes video structuring tasks, physical sign extraction tasks, target appearance comparison tasks, etc. Global target recognition extracts all target data through video structuring, and then extracts target appearance features from the target data. After the target appearance features are extracted, the target appearance is compared with a preset feature library. If it is not in the preset feature library, it is marked as a suspicious target.

[0028] The plan can also link different analysis functions. For example, the behavior analysis plan can include general behavior analysis and three-dimensional behavior analysis, and the environmental detection plan can include sound detection tasks and object monitoring tasks.

[0029] There may be one or more analysis plans associated with a target event. Therefore, when a target event is detected, the video data may be analyzed from multiple angles to obtain a more comprehensive analysis result.

[0030] The analysis tasks representing the advanced analysis included in the analysis plan can be allocated as needed, and this application does not impose any restrictions; the analysis plan associated with the target event can also be allocated as needed, and this application does not impose any restrictions.

[0031] The above scheme first performs basic analysis based on video data to obtain basic analysis results. When the basic analysis results indicate that a target event has been detected, the associated analysis plan is called to perform advanced analysis. Since advanced analysis is performed only when a target event is detected, it will not occupy more computing resources when no target event is detected (i.e., when advanced analysis is not needed), thereby saving computing resources. At the same time, when a target event is detected, the associated analysis plan can be called to perform advanced analysis to obtain advanced analysis results. Further analysis results can be obtained through advanced analysis when needed, which can reduce missed reports and false alarms.

[0032] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.

[0033] In some embodiments, performing basic analysis based on video data to obtain basic analysis results includes: Step S21: Perform basic analysis based on the video data of the current video channel to obtain a basic analysis result of the current video channel.

[0034] In response to the basic analysis result indicating that a target event is detected, an analysis plan associated with the target event is called, and an advanced analysis is performed on the video data based on the analysis plan, and the advanced analysis results obtained include: Step S41: In response to the basic analysis result of the current video channel indicating that a target event is detected, an analysis plan associated with the target event is called, and an advanced analysis is performed on the video data of the current video channel based on the analysis plan to obtain an advanced analysis result.

[0035] In some embodiments, performing basic analysis based on video data to obtain basic analysis results includes: Step S21: Perform basic analysis based on the video data of the current video channel to obtain a basic analysis result of the current video channel.

[0036] In response to the basic analysis result indicating that a target event is detected, an analysis plan associated with the target event is called, and an advanced analysis is performed on the video data based on the analysis plan, and the advanced analysis results obtained include: Step S41: In response to the basic analysis result of the current video channel indicating that a target event has been detected, an analysis plan associated with the target event is called, and based on the analysis plan, an advanced analysis is performed on the video data of the current video channel and the associated channels of the current video channel to obtain an advanced analysis result; wherein the associated channels of the current video channel include video channels surrounding the current video channel.

[0037] As mentioned above, the video channel can provide video data streams shot by different camera devices. Different camera devices can shoot different areas, thereby obtaining camera images of different areas. Different areas can be correlated in geographical location. For example, area A and area B can be adjacent or overlapped in geographical location, or area B can be reached from area A via a certain reachable path, etc. Therefore, the shooting area corresponding to the current video channel can be correlated in geographical location with the shooting area corresponding to its associated channel, that is, the associated channel of the current video channel includes the video channels around the current video channel.

[0038] When the basic analysis result obtained by performing basic analysis on the video data of the current video channel indicates that a target event has been detected, the associated analysis plan is called to perform advanced analysis on the video data of the associated channel, which can combine the camera images of multiple areas to obtain more comprehensive analysis results. For example, when performing tripwire intrusion detection based on the video data of area A, it is detected that a target object has crossed the dividing line (that is, the target event has been detected). In addition to performing advanced analysis on the video data of area A based on the analysis plan, advanced analysis is also performed on areas B and C (areas B and C are geographically related to area A) based on the analysis plan, which can not only obtain more video data for analysis to obtain more information, but also trace the movement of the target object.

[0039] In some embodiments, performing basic analysis based on video data to obtain basic analysis results includes: Step S21: Perform basic analysis based on the video data of the current video channel to obtain basic analysis results and the moving direction of the target object that triggers the target event.

[0040] In response to the basic analysis result of the current video channel indicating that a target event is detected, an analysis plan associated with the target event is called, and based on the analysis plan, an advanced analysis is performed on the video data of the current video channel and the associated channels of the current video channel, and before obtaining the advanced analysis result, the method further includes: Step S31: Determine the associated channel of the current video channel based on the moving direction of the target object.

[0041] In the aforementioned embodiment, the associated channel of the current video channel refers to the corresponding shooting area having a correlation in geographical location with the shooting area corresponding to the current video channel. In this embodiment, further according to the moving direction of the target object, the video channel corresponding to the moving direction of the target object is screened out from the video channels having a correlation in geographical location with the shooting area corresponding to the current video channel. Exemplarily, the shooting area corresponding to the current video channel is area A, and the shooting areas having a correlation in geographical location with area A include area B and area C. According to the basic analysis, the moving direction of the target object is obtained toward area C, and the video channel corresponding to area C is determined to be the associated channel of the current video channel. Therefore, determining the associated channel based on the moving direction of the target object and then performing advanced analysis on the video data of the associated channel can reduce the amount of video data for advanced analysis and further save computing resources.

[0042] In some embodiments, in response to the basic analysis result indicating that a target event is detected, an analysis plan associated with the target event is called, and an advanced analysis is performed on the video data based on the analysis plan, and the advanced analysis result obtained includes: Step S41: In response to the basic analysis result indicating that a first target event is detected, a first analysis plan associated with the first target event is called, and a first advanced analysis is performed on the video data based on the first analysis plan to obtain a first advanced analysis result.

[0043] Step S42: In response to the first advanced analysis result indicating that a second target event is detected, a second analysis plan associated with the second target event is called, and a second advanced analysis is performed on the video data based on the second analysis plan to obtain a second advanced analysis result.

[0044] The first analysis plan includes at least one first advanced analysis task, and the second analysis plan includes at least one second advanced analysis task.

[0045] Exemplarily, assuming that the basic analysis includes line intrusion detection, when the target object is detected crossing the demarcation line (i.e., the first target event) during line intrusion detection, the first analysis plan associated with the first target event, such as the target appearance analysis plan, is called. Based on the target appearance analysis plan, the first advanced analysis is performed on the video data to obtain the first advanced analysis result.

[0046] The first analysis plan associated with the first target event may be associated based on a predetermined condition or based on a predetermined rule. The predetermined condition may be that the basic analysis result indicates that the first target event is detected. The predetermined rule may be a rule pre-set according to actual needs. For example, if the basic analysis result of the current video channel indicates that the first target event is detected, then the associated channel of the current video channel is further detected based on the target appearance analysis plan.

[0047] In some embodiments, performing a first advanced analysis on the video data based on a first analysis plan to obtain a first advanced analysis result includes: performing a first advanced analysis on the video data according to the first analysis plan based on the video data and the basic analysis result to obtain a first advanced analysis result.

[0048] Performing a second advanced analysis on the video data based on the second analysis plan to obtain a second advanced analysis result includes: performing a second advanced analysis on the video data according to the second analysis plan based on the video data, the basic analysis result and the first advanced analysis result to obtain a second advanced analysis result.

[0049] That is to say, when the current plan or the previous analysis function triggers a new plan or a new analysis function, the previous plan or the previous analysis function can pass the analysis results obtained by the previous plan or the previous analysis function to the new plan or the new analysis function, so that the new plan or the new analysis function can perform further analysis based on the video data and the analysis results of the previous plan or the previous analysis function, thereby being able to obtain more accurate video information based on more data.

[0050] In some embodiments, please refer to Figure 2 The video data analysis method provided in this application can be used according to Figure 2 The process shown is carried out.

[0051] Assuming that a basic analysis (such as tripwire intrusion detection) is performed based on the video data stream of video channel 1, and a tripwire intrusion event is detected, the analysis plan associated with the tripwire intrusion event is called for advanced analysis. Figure 2 As shown, the advanced analysis tasks in the analysis plan associated with the tripwire intrusion event include target feature extraction. The detection of the tripwire intrusion event for video channel 1 can further trigger the target appearance recognition for video channel 2. Furthermore, the analysis result of the target appearance recognition for video channel 2 can trigger the suspicious target group analysis plan when it meets the predetermined conditions. For example, when the target appearance obtained by the target appearance recognition does not match any appearance feature in the predetermined appearance library, the suspicious target group analysis plan is triggered. After the advanced analysis based on the target group analysis plan, the behavior analysis monitoring and environmental monitoring plans can be further triggered when the analysis results meet certain conditions. Finally, the behavior analysis monitoring and environmental monitoring plans will pass all the analysis results and data obtained to the event nature determination plan. And, in Figure 2In the process, the previous plan or the previous analysis function can pass the analysis results obtained by the previous plan or the previous analysis function to the new plan or the new analysis function. For example, after performing advanced analysis based on the target appearance analysis plan for video channel 2, the advanced analysis results obtained by the advanced analysis based on the target appearance analysis plan and the basic analysis results obtained by performing tripwire intrusion detection on video channel 1 can also be passed to the next stage, that is, the suspicious target group analysis plan.

[0052] Therefore, the target events generated by basic analysis drive advanced analysis, and advanced analysis can drive other types of advanced analysis to obtain more information. Finally, these target events and information are fully analyzed. In this way, more relevant information can be obtained when an event occurs, and the target event can be reported more accurately. Users can also define various plans according to established rules, and combine various plans into special plans for a certain scenario as needed. In this way, when an event occurs, the necessary information can be collected in a timely manner and the cause of the event can be preliminarily analyzed, providing effective information for subsequent tracing.

[0053] In some embodiments, the method further includes, in response to calling an analysis plan associated with the target event, scheduling computing resources required for advanced analysis tasks in the analysis plan to perform advanced analysis.

[0054] As in the aforementioned embodiment, different analysis tasks require different computing resources. Therefore, scheduling and analyzing the computing resources required for advanced analysis tasks in the plan can further reduce the waste of computing resources.

[0055] In some embodiments, in response to calling an analysis plan corresponding to a target event, and scheduling the computing power resources required for advanced analysis tasks in the analysis plan to perform advanced analysis, obtaining advanced analysis results includes: in response to calling an analysis plan corresponding to a target event, adding the analysis plan to a queue to be processed; sorting each analysis plan in the queue to be processed according to the priority of each analysis plan in the queue to be processed; and scheduling the computing power resources required for advanced analysis tasks in each analysis plan in the queue to be processed in turn to perform advanced analysis.

[0056] In some embodiments, the basic analysis function can be enabled for all video channels; the advanced analysis can be enabled in accordance with the basic analysis function. For example, the basic analysis function of intelligent motion detection, which can roughly distinguish between people and vehicles, is a basic-level intelligent capability, and the corresponding advanced analysis capability is structured analysis, which is an intelligent analysis function that can distinguish the characteristics of people and vehicles; for example, when target appearance detection is used as a basic analysis function, the corresponding advanced analysis is to conduct further intelligent analysis on the characteristics of the target appearance or target shape, such as target appearance comparison, target shape detection, and body shape detection; so the deployment of advanced analysis usually corresponds to the basic analysis, and the advanced analysis is not for specific video channels, but for all corresponding video channels with basic analysis enabled; at the same time, the advanced analysis is not run in real time, and it will only be started after the corresponding basic analysis has the corresponding analysis results. In addition, the advanced analysis can also be deployed on the basis of other advanced analyses.

[0057] In some embodiments, in terms of computing resource scheduling, since basic analysis needs to cover almost all video channels, it is generally bound to the video channels, so this basic analysis usually does not need to be scheduled.

[0058] At the advanced analysis level, there are several calling methods: queue scheduling, priority scheduling, custom scheduling and intelligent scheduling.

[0059] Queuing scheduling is a scheduling method that performs advanced analysis in the order in which the results of basic analysis are generated.

[0060] Priority scheduling is to prioritize basic analysis according to its intrinsic weight, and process advanced analysis within a certain period of time according to priority. For example, if the weight of tripwire intrusion is greater than that of intelligent motion detection, then the priority of tripwire intrusion is greater than that of intelligent motion detection.

[0061] User-defined scheduling: Users can customize scheduling by adjusting the priority of basic analysis or specifying the priority of an advanced analysis.

[0062] Intelligent scheduling optimizes and adjusts priorities in real time by analyzing past results and user behavior information. For example, it statistically analyzes the final results after full analysis and records basic analysis of the sources of these results, or statistically analyzes the results of user browsing intelligent analysis and records basic analysis of user operations such as number of views and operations (such as backup, favorites, and tagging). Finally, it performs weighted analysis on these analysis results and appropriately adjusts the priority of the corresponding basic analysis.

[0063] In some embodiments, the computing power resource may be an analyzer for processing various analysis tasks, for example, a target appearance analyzer for target appearance analysis, etc.

[0064] In actual deployment, each scheduler can schedule all analyzers, and the scheduler is not limited to the current device, but can also be other idle devices with the same capabilities in the network. Therefore, a scheduling management system is needed to manage all current schedulers, and the scheduler must also report the status of its work.

[0065] Therefore, the present application also provides a scheduling management system, referring to Figure 3 The scheduling management system 300 may include a target appearance scheduler 310 and a target shape scheduler 320, which are used to schedule the target appearance analyzer and the target shape analyzer respectively. Of course, the scheduling management system 300 may also include other schedulers. For the sake of convenience, Figure 3 It is not shown in the figure, and only the target appearance scheduler 310 and the target shape scheduler 320 are used as examples for explanation. The target appearance scheduler 310 can schedule the local target appearance analyzer 311 and the local target appearance analyzer 312 so that each local target appearance analyzer performs target appearance analysis on the video data stream of each video channel, and can also schedule the host target appearance analyzer 313 of the network host. The target shape scheduler 320 can schedule the local target shape analyzer 321 and the local target shape analyzer 322 so that each local target shape analyzer performs target shape analysis on the video data stream of each video channel, and can also schedule the host target shape analyzer 323 of the network host. Among them, the local target shape analyzer and the local target appearance analyzer refer to the analyzers locally possessed by the device, and the analyzer of the network host refers to the analyzer of other devices in the network.

[0066] The scheduling management system can be deployed separately in the device to manage the local scheduler of the device. When it is necessary to manage scheduler resources between different devices, the computing power resources between various devices in the network system can be realized by setting the master-slave configuration of the scheduling management system. The specific strategy of the master-slave configuration is as follows: During deployment, a certain device can be set as the host of the scheduling management system, and other devices can be set as slaves. The slave has the IP address of the host. The slave will register itself to the host, that is, the host also has the IP addresses and related information of all slaves. When the status of the local analyzer resources (i.e. computing power resources) managed by the scheduler of the slave changes, it needs to be reported to the host. After receiving the reported information, the host will immediately summarize and update the resource status of all devices in the network, and then send the latest status to the slave. There is a keep-alive mechanism between the master and slave devices. When the host device cannot be affected for some reason, a new host device can be selected through voting between the machine devices. The voting mainly considers network conditions and device performance. After the new host device is determined, it will take over the task of the scheduling management system from the original host.

[0067] See also Figure 4 , Figure 4 It is a schematic diagram of a framework of an embodiment of a video data analysis device of the present application. The video data analysis device 40 includes a basic analysis module 41 and an advanced analysis module 42; the basic analysis module 41 is used to perform basic analysis based on video data to obtain basic analysis results; the advanced analysis module 42 is used to respond to the basic analysis result indicating that a target event has been detected, call an analysis plan associated with the target event, and perform advanced analysis on the video data based on the analysis plan to obtain an advanced analysis result; wherein the analysis plan includes at least one analysis task representing advanced analysis, and the computing power resources required for basic analysis are less than the computing power resources required for advanced analysis.

[0068] The above scheme first performs basic analysis based on video data to obtain basic analysis results. When the basic analysis results indicate that a target event has been detected, the associated analysis plan is called to perform advanced analysis. Since advanced analysis is performed only when a target event is detected, it will not occupy more computing resources when no target event is detected (i.e., when advanced analysis is not needed), thereby saving computing resources. At the same time, when a target event is detected, the associated analysis plan can be called to perform advanced analysis to obtain advanced analysis results. Further analysis results can be obtained through advanced analysis when needed, which can reduce missed reports and false alarms.

[0069] The basic analysis module 41 and the advanced analysis module 42 may be specifically described in the embodiments of the video data analysis method, which will not be described in detail here.

[0070] In some embodiments, the basic analysis module 41 includes a current video channel basic analysis unit, which is used to perform basic analysis based on the video data of the current video channel to obtain a basic analysis result of the current video channel. The advanced analysis module 42 includes a current video channel basic analysis unit, which is used to call an analysis plan associated with the target event in response to the basic analysis result of the current video channel indicating that a target event has been detected, and perform advanced analysis on the video data of the current video channel based on the analysis plan to obtain an advanced analysis result.

[0071] In some embodiments, the basic analysis module 41 includes an associated video channel basic analysis unit, which is used to perform basic analysis based on the video data of the associated video channel to obtain a basic analysis result of the current video channel. The advanced analysis module 42 includes an associated video channel basic analysis unit, which is used to call an analysis plan associated with the target event in response to the basic analysis result of the associated video channel indicating that a target event has been detected, and perform advanced analysis on the video data of the associated video channel based on the analysis plan to obtain an advanced analysis result.

[0072] In some embodiments, the basic analysis module 41 includes a moving direction analysis unit for performing basic analysis based on the video data of the current video channel to obtain a basic analysis result and a moving direction of the target object that triggers the target event. The video data analysis device 40 also includes an associated channel determination module for determining an associated channel of the current video channel based on the moving direction of the target object.

[0073] In some embodiments, the advanced analysis module 42 includes an advanced analysis unit for, in response to the basic analysis result indicating the detection of a first target event, calling a first analysis plan associated with the first target event, performing a first advanced analysis on the video data based on the first analysis plan, and obtaining a first advanced analysis result; in response to the first advanced analysis result indicating the detection of a second target event, calling a second analysis plan associated with the second target event, performing a second advanced analysis on the video data based on the second analysis plan, and obtaining a second advanced analysis result; wherein the first analysis plan includes at least one first advanced analysis task, and the second analysis plan includes at least one second advanced analysis task.

[0074] In some embodiments, the advanced analysis unit includes a first advanced analysis unit and a second advanced analysis unit. The first advanced analysis unit is used to perform a first advanced analysis based on the video data according to a first analysis plan based on the video data and the basic analysis results to obtain a first advanced analysis result; the second advanced analysis unit is used to perform a second advanced analysis based on the video data according to a second analysis plan based on the video data, the basic analysis results and the first advanced analysis results to obtain a second advanced analysis result.

[0075] See also Figure 5 , Figure 5 : is a schematic diagram of a framework of an embodiment of an electronic device of the present application. The electronic device 50 includes a memory 51 and a processor 52 coupled to each other, and the processor 52 is used to execute program instructions stored in the memory 51 to implement the steps of any of the above-mentioned video data analysis method embodiments. In a specific implementation scenario, the electronic device 50 may include but is not limited to: a microcomputer, a server, and in addition, the electronic device 50 may also include a mobile device such as a laptop computer and a tablet computer, which is not limited here.

[0076] Specifically, the processor 52 is used to control itself and the memory 51 to implement the steps in any of the above-mentioned video data analysis method embodiments. The processor 52 can also be called a CPU (Central Processing Unit). The processor 52 may be an integrated circuit chip with signal processing capabilities. The processor 52 can also be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. In addition, the processor 52 can be implemented by an integrated circuit chip.

[0077] The above scheme first performs basic analysis based on video data to obtain basic analysis results. When the basic analysis results indicate that a target event has been detected, the associated analysis plan is called to perform advanced analysis. Since advanced analysis is performed only when a target event is detected, it will not occupy more computing resources when no target event is detected (i.e., when advanced analysis is not needed), thereby saving computing resources. At the same time, when a target event is detected, the associated analysis plan can be called to perform advanced analysis to obtain advanced analysis results. Further analysis results can be obtained through advanced analysis when needed, which can reduce missed reports and false alarms.

[0078] See also Figure 6 , Figure 6 The computer-readable storage medium 60 stores program instructions 601 that can be executed by a processor, and the program instructions 601 are used to implement the steps in any of the above-mentioned video data analysis method embodiments.

[0079] The above scheme first performs basic analysis based on video data to obtain basic analysis results. When the basic analysis results indicate that a target event has been detected, the associated analysis plan is called to perform advanced analysis. Since advanced analysis is performed only when a target event is detected, it will not occupy more computing resources when no target event is detected (i.e., when advanced analysis is not needed), thereby saving computing resources. At the same time, when a target event is detected, the associated analysis plan can be called to perform advanced analysis to obtain advanced analysis results. Further analysis results can be obtained through advanced analysis when needed, which can reduce missed reports and false alarms.

[0080] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0081] The above description of various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced to each other, and for the sake of brevity, they will not be repeated herein.

[0082] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation, such as 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, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0083] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0084] 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 technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of each implementation method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

[0085] If the technical solution of this application involves personal information, the product using the technical solution of this application has clearly informed the personal information processing rules and obtained the individual's voluntary consent before processing the personal information. If the technical solution of this application involves sensitive personal information, the product using the technical solution of this application has obtained the individual's separate consent before processing the sensitive personal information, and at the same time meets the "explicit consent" requirement. For example, on personal information collection devices such as cameras, clear and prominent signs are set to inform that the personal information collection scope has been entered and personal information will be collected. If the individual voluntarily enters the collection scope, it is deemed that he or she agrees to the collection of his or her personal information; or on the device that processes personal information, the personal information processing rules are notified by obvious signs / information, and the individual's authorization is obtained through pop-up information or by asking the individual to upload his or her personal information; among them, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the type of personal information processed.

Claims

1. A video data analysis method, characterized in that: include: Perform basic analysis based on video data to obtain basic analysis results; In response to the basic analysis result indicating that a target event is detected, calling an analysis plan associated with the target event, and performing an advanced analysis on the video data based on the analysis plan to obtain an advanced analysis result; The analysis plan includes at least one analysis task that characterizes the advanced analysis, and the computing power resources required for the basic analysis are less than the computing power resources required for the advanced analysis.

2. The video data analysis method according to claim 1, characterized in that: The basic analysis based on the video data is performed to obtain the basic analysis results including: Performing basic analysis based on the video data of the current video channel to obtain the basic analysis result of the current video channel; In response to the basic analysis result indicating that a target event is detected, calling an analysis plan associated with the target event, and performing advanced analysis on the video data based on the analysis plan to obtain an advanced analysis result includes: In response to the basic analysis result of the current video channel indicating that a target event is detected, an analysis plan associated with the target event is called, and an advanced analysis is performed on the video data of the current video channel based on the analysis plan to obtain an advanced analysis result.

3. The video data analysis method according to claim 1, characterized in that: The basic analysis based on the video data is performed to obtain the basic analysis results including: Performing basic analysis based on the video data of the current video channel to obtain the basic analysis result of the current video channel; In response to the basic analysis result indicating that a target event is detected, calling an analysis plan associated with the target event, and performing advanced analysis on the video data based on the analysis plan to obtain an advanced analysis result includes: In response to the basic analysis result of the current video channel indicating that a target event has been detected, an analysis plan associated with the target event is called, and based on the analysis plan, an advanced analysis is performed on the video data of the current video channel and the associated channels of the current video channel to obtain an advanced analysis result; wherein the associated channels of the current video channel include video channels surrounding the current video channel.

4. The video data analysis method according to claim 3, characterized in that: The basic analysis based on the video data is performed to obtain the basic analysis results including: Perform basic analysis based on the video data of the current video channel to obtain basic analysis results and the moving direction of the target object that triggers the target event; In response to the basic analysis result of the current video channel indicating that a target event is detected, an analysis plan associated with the target event is called, and based on the analysis plan, an advanced analysis is performed on the video data of the current video channel and the associated channels of the current video channel, and before obtaining the advanced analysis result, the method further includes: Based on the moving direction of the target object, an associated channel of the current video channel is determined.

5. The video data analysis method according to any one of claims 1 to 4, characterized in that: In response to the basic analysis result indicating that a target event is detected, calling an analysis plan associated with the target event, and performing advanced analysis on the video data based on the analysis plan to obtain an advanced analysis result includes: In response to the basic analysis result indicating that a first target event is detected, calling a first analysis plan associated with the first target event, performing a first advanced analysis on the video data based on the first analysis plan, and obtaining a first advanced analysis result; In response to the first advanced analysis result indicating that a second target event is detected, calling a second analysis plan associated with the second target event, performing a second advanced analysis on the video data based on the second analysis plan, and obtaining a second advanced analysis result; The first analysis plan includes at least one first advanced analysis task, and the second analysis plan includes at least one second advanced analysis task.

6. The video data analysis method according to claim 5, characterized in that: The performing the first advanced analysis on the video data based on the first analysis plan to obtain the first advanced analysis result includes: Based on the video data and the basic analysis result, performing the first advanced analysis based on the video data according to the first analysis plan to obtain a first advanced analysis result; The performing the second advanced analysis on the video data based on the second analysis plan to obtain the second advanced analysis result includes: Based on the video data, the basic analysis result and the first advanced analysis result, the second advanced analysis is performed based on the video data according to the second analysis plan to obtain a second advanced analysis result.

7. The video data analysis method according to claim 1, characterized in that: Also includes: In response to calling the analysis plan associated with the target event, the computing resources required for the advanced analysis task in the analysis plan are scheduled to perform the advanced analysis.

8. The video data analysis method according to claim 7, characterized in that: The step of responding to calling the analysis plan corresponding to the target event and scheduling the computing resources required by the advanced analysis task in the analysis plan to perform advanced analysis to obtain the advanced analysis result includes: In response to calling the analysis plan corresponding to the target event, adding the analysis plan to a queue to be processed; Sorting the analysis plans in the queue to be processed according to their priorities; The computing resources required for the advanced analysis tasks in each analysis plan in the queue to be processed are scheduled in sequence for advanced analysis.

9. An electronic device, characterized in that: It comprises a memory and a processor coupled to each other, wherein the processor is used to execute program instructions stored in the memory, and the program instructions are executed by the processor to implement the video data analysis method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: Program instructions are stored thereon, and the program instructions are executed by a processor to implement the video data analysis method according to any one of claims 1 to 8.

Images