Digital video device control method and device

Abstract

The invention relates to the technical field of digital video devices, in particular to a digital video device control method and device. The method comprises the following steps: acquiring digital monitoring data, performing real-time scene acquisition and dynamic region optimization according to the data, and generating optimized dynamic region data; video frames are read based on the dynamic region optimization data, and abnormal behavior analysis and behavior intention evaluation are carried out; calculating the complexity of monitoring behaviors by combining video frame data and behavior intention data, and evaluating resource consumption increase; the processing bottleneck is evaluated according to the resource consumption growth data, and the processing capability attenuation analysis is carried out, so that the monitoring device is adjusted, and the system performance is optimized; according to the method, the resource allocation and processing capability of the video device is dynamically adjusted through real-time analysis and processing of monitoring data. According to the invention, the digital video device is adjusted, so that the prediction identification of the digital video device is more accurate and efficient.

Description

Technical Field

[0001] The present invention relates to the technical field of digital video devices, and in particular to a digital video device control method and device. Background Art

[0002] Traditional analog video surveillance systems are gradually being replaced by intelligent surveillance systems based on digital video devices due to limitations in image quality, transmission distance, data storage and processing capabilities. Digital video devices have significant advantages such as high definition, remote control, flexible data storage, and intelligent data processing, and have become the core equipment in modern surveillance systems. Digital video devices mainly include several core parts such as video acquisition, processing, transmission, storage and display. In the video acquisition stage, high-definition cameras (such as network cameras) are responsible for acquiring video signals of the surveillance scene and converting analog signals into digital signals through digital processing. Traditional video surveillance systems can no longer meet the requirements of high efficiency, intelligence, and real-time. For example, in places such as shopping malls and stations, how to timely grasp the changes in crowd density, discover abnormal behaviors, and warn of potential threats, however, traditional digital video devices have the problem of inaccurate prediction of surveillance behaviors and inaccurate prediction of surveillance behavior processing bottlenecks. Summary of the invention

[0003] Based on this, it is necessary to provide a digital video device control method and device to solve at least one of the above technical problems.

[0004] To achieve the above object, a digital video device control method comprises the following steps:

[0005] Step S1: acquiring digital monitoring data; acquiring real-time monitoring scenes according to the digital monitoring data; obtaining real-time monitoring scene data; optimizing the dynamic area of ​​the monitoring video according to the real-time monitoring scene data, and obtaining dynamic area optimization data of the monitoring video;

[0006] Step S2: Read the dynamic area video frame according to the dynamic area optimization data of the surveillance video to obtain the video dynamic area video frame data; perform dynamic area abnormal behavior analysis according to the video dynamic area video frame data to obtain the dynamic area abnormal behavior data; perform dynamic area behavior intention evaluation according to the dynamic area abnormal behavior data to obtain the dynamic area behavior intention data;

[0007] Step S3: Calculate the complexity of video monitoring behavior according to the video frame data of the dynamic area of ​​the video and the behavior intention data of the dynamic area to obtain the complexity data of video monitoring behavior; evaluate the growth of monitoring video resource consumption according to the complexity data of video monitoring behavior to obtain the growth data of monitoring video resource consumption;

[0008] Step S4: perform monitoring behavior processing bottleneck assessment based on monitoring video resource consumption growth data to obtain monitoring behavior processing bottleneck data; perform monitoring behavior processing capacity attenuation analysis on the monitoring behavior processing bottleneck data to obtain monitoring behavior processing capacity attenuation data; adjust the digital monitoring video device based on the monitoring behavior processing capacity attenuation data to obtain digital monitoring device adjustment data.

[0009] The present invention obtains digital monitoring data and performs real-time scene acquisition, which can ensure that the system can make timely responses based on the current monitoring environment and data dynamics. Through the capture and analysis of real-time scene data, the monitoring system can efficiently identify the areas that need to be focused on, and then optimize the dynamic areas of the monitoring video. This dynamic area optimization helps the system to concentrate resources to process data in key areas, without having to process indiscriminately in all areas, thereby improving processing efficiency and resource utilization. When further processing the video dynamic area frame data, abnormal behavior analysis can effectively detect potential abnormal or intrusion behaviors, ensuring that the system can make timely responses when emergencies occur. Through subsequent analysis and evaluation of these abnormal behaviors, the system can not only identify the behavior pattern, but also infer the intention behind the behavior, which is of great value for the triggering of the early warning mechanism and the decision support of security. The evaluation of the dynamic area behavior intention enables the system to judge the threat level of the behavior, thereby dynamically adjusting the monitoring strategy and response priority. By combining the video dynamic area video frame data and the behavior intention data, the system can calculate the complexity of video monitoring behavior, which is an important indicator for evaluating the monitoring data processing load. By calculating the complexity of monitoring behavior, the system can better understand the processing requirements of data, predict the processing pressure of the system, and avoid processing bottlenecks in high-complexity scenarios. As the complexity of behavior increases, the monitoring system will adjust according to the growth trend of resource consumption, and ensure that the system is always within the controllable range by timely evaluating the growth of resource consumption. Under high load conditions, the evaluation of the monitoring behavior processing bottleneck can identify the bottleneck of the system processing capacity and help system managers understand the performance problems encountered in specific situations. Next, the processing capacity decay analysis is performed to further reveal the performance decay phenomenon that occurs when the processing pressure is too large, and predict the system's extreme processing capacity. Through this step, the monitoring system can make timely adjustments when the load is too high, avoid system crashes or response delays, and ensure real-time performance and stability. According to the evaluation results of the monitoring behavior processing capacity decay, the system can adjust the digital monitoring video device, optimize the hardware and software configuration according to the actual resources and processing requirements, so as to ensure that it can continue to operate efficiently under high load conditions. Therefore, the present invention is an optimization process for a traditional digital video device control method. The traditional digital video device has the problem of inaccurate prediction of monitoring behavior and inaccurate prediction of monitoring behavior processing bottlenecks. The accuracy of monitoring behavior prediction and monitoring behavior processing bottleneck prediction has been improved.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: Acquire digital surveillance video data;

[0012] Step S12: acquiring the real-time scene of the monitoring video according to the digital monitoring video data; obtaining the real-time scene data of the monitoring video;

[0013] Step S13: collecting low-resolution monitoring area data according to the real-time scene data of the monitoring video to obtain low-resolution video monitoring area data;

[0014] Step S14: optimizing the dynamic region of the surveillance video according to the low-resolution region data of the video surveillance to obtain optimized dynamic region data of the surveillance video.

[0015] By acquiring digital surveillance video data, the present invention enables the system to capture the latest dynamics of the surveillance scene in real time, thereby providing basic data for subsequent analysis and decision-making. On this basis, the acquisition of real-time scene data ensures that the monitoring system can keep up with environmental changes, identify and respond to potential risks in a timely manner. The acquisition of low-resolution areas optimizes resource allocation, allowing the monitoring system to focus on more critical areas without wasting computing resources on non-critical areas. This step not only improves data processing efficiency, but also effectively reduces storage requirements and computing loads. Further dynamic area optimization improves the real-time and accuracy of monitoring by accurately demarcating high-priority areas that need attention, reduces redundant data analysis, and ensures that the system can still maintain efficient operation under limited resources. The beneficial effects of this series of steps are reflected in improving monitoring efficiency, accuracy and response speed, while playing an important role in maintaining system stability and resource optimization.

[0016] Preferably, step S14 comprises the following steps:

[0017] Step S141: performing video grayscale image conversion according to the video surveillance low-resolution area data to obtain video area grayscale image data;

[0018] Step S142: performing dynamic region acquisition according to the video region grayscale image data to obtain video region grayscale image dynamic region data;

[0019] Step S143: performing dynamic region edge detection according to the video grayscale image dynamic region data to obtain dynamic region edge data;

[0020] Step S144: adjusting the edge detail resolution of the dynamic area edge data to obtain dynamic area edge detail adjusted data;

[0021] Step S145: performing foreground object separation according to the video grayscale image dynamic area data to obtain grayscale image foreground object separation data;

[0022] Step S146: Optimize the dynamic area of ​​the surveillance video based on the grayscale image foreground object separation data and the dynamic area edge detail adjustment data to obtain surveillance video dynamic area optimization data.

[0023] The present invention converts the low-resolution area data of video surveillance into grayscale image data, so that the system can simplify the processing process and improve the analysis efficiency, especially in the case of limited resources, effectively reducing the computational burden. Based on the grayscale image data, dynamic area acquisition can focus on the key area, accurately identify the important changes in the surveillance video, and improve the accuracy of the analysis. On this basis, the dynamic area edge detection enables the system to identify the key boundaries of the area, provide clearer visual information, and facilitate subsequent behavior analysis and target recognition. Further edge detail resolution adjustment optimizes the clarity of the edge image, enhances the detail expression of the dynamic area, and ensures the adaptability of the system to complex scenes. Through foreground object separation, the system can effectively distinguish the key targets in the video, reduce background interference, and provide higher quality data for behavior analysis. Combining foreground object separation data and edge detail adjustment data, it can optimize the monitoring effect of the dynamic area, improve the processing accuracy and response speed of the surveillance video, and achieve efficient and accurate monitoring data analysis and resource optimization.

[0024] Preferably, step S2 comprises the following steps:

[0025] Step S21: Read the dynamic area video frame according to the dynamic area optimization data of the monitoring video to obtain the video dynamic area video frame data;

[0026] Step S22: performing behavior recognition in the dynamic region on the video frame data of the video dynamic region to obtain behavior data of the video dynamic region;

[0027] Step S23: Performing dynamic area safety behavior analysis based on the video dynamic area behavior data to obtain dynamic area safety behavior data;

[0028] Step S24: performing dynamic area intrusion behavior analysis according to the video dynamic area behavior data to obtain dynamic area intrusion behavior data;

[0029] Step S25: performing dynamic area abnormal behavior analysis on the dynamic area intrusion behavior data and the dynamic area security behavior data to obtain dynamic area abnormal behavior data;

[0030] Step S26: Perform dynamic area behavior intention assessment based on the dynamic area abnormal behavior data to obtain dynamic area behavior intention data.

[0031] The present invention reads video frames by optimizing the data of dynamic area of ​​monitoring video, and the system can accurately extract real-time data of key areas, providing a high-quality information basis for subsequent behavior analysis. Behavior recognition in dynamic areas ensures that the system can effectively identify and distinguish different behavior patterns, further improving the intelligent level of monitoring. Security behavior analysis helps the system to actively identify normal behavior, provides a reference basis for the determination of potential threats, and improves the security of overall monitoring. On this basis, intrusion behavior analysis can accurately identify abnormal activities that do not conform to the normal pattern, timely warn and take necessary countermeasures. By comprehensively analyzing intrusion behavior and security behavior, dynamic area abnormal behavior analysis further improves the system's resilience, enabling it to identify and respond to a variety of emergencies. Behavior intention assessment combined with dynamic area abnormal behavior data can analyze the potential intentions behind the behavior, providing a scientific basis for event prediction and decision support. Through this series of steps, the system can achieve accurate monitoring and efficient response to the monitoring area, effectively improving the safety and emergency response capabilities of the monitoring site.

[0032] Preferably, step S23 includes the following steps:

[0033] Step S231: collecting motion posture data according to the video dynamic area behavior data to obtain motion behavior posture data;

[0034] Step S232: Calculating the movement inclination according to the movement behavior posture data to obtain movement posture inclination data;

[0035] Step S233: performing a motion speed reduction calculation on the motion behavior posture data to obtain motion speed reduction data;

[0036] Step S234: performing a motion balance attenuation evaluation according to the motion action speed deceleration data and the motion posture inclination data to obtain motion balance attenuation data;

[0037] Step S235: Calculating the fall probability based on the sports balance attenuation data and the sports action speed reduction data to obtain sports fall probability data;

[0038] Step S236: Perform dynamic area safety behavior analysis based on the motion speed reduction data and the motion fall probability data to obtain dynamic area safety behavior data.

[0039] The present invention can accurately capture the individual behavior patterns in motion through motion posture collection, and provide detailed dynamic information for subsequent behavior analysis. The motion inclination calculation helps to identify the individual's posture changes, especially during walking or exercise. By analyzing the inclination data, it can be timely determined whether there is an abnormal motion posture, and then infer the potential risk of falling or imbalance. The calculation of the slowing down of the movement speed can reveal abnormal changes in the movement, such as sudden stagnation or deceleration, indicating health problems or unexpected behaviors. Further motion balance attenuation evaluation helps to judge the stability of the individual, especially for the elderly or physically weak individuals, and can predict and prevent accidents such as falls in advance. Combined with the calculation of the probability of falling, the system can generate an accurate fall risk assessment based on real-time data, trigger the early warning mechanism in time, reduce the probability of accidents, and use the speed reduction data and the probability of falling data for dynamic regional safety behavior analysis, which can comprehensively evaluate the safety and potential risks in the area, thereby improving the accuracy and reaction speed of the monitoring system in handling emergencies, and ensuring that safety is effectively guaranteed.

[0040] Preferably, step S24 includes the following steps:

[0041] Step S241: extracting the climbing behavior according to the video dynamic area behavior data to obtain the video dynamic area climbing behavior data;

[0042] Step S242: dividing the video dynamic area climbing behavior data into climbing prohibited area behaviors to obtain climbing prohibited area behavior data;

[0043] Step S243: evaluating the intrusion behavior line according to the video dynamic area behavior data and the forbidden area crossing behavior data to obtain the intrusion behavior line data;

[0044] Step S244: performing intrusion behavior destination analysis based on the intrusion behavior route data to obtain intrusion behavior destination data;

[0045] Step S245: Perform dynamic area intrusion behavior analysis based on the intrusion behavior destination data and the prohibited area crossing behavior data to obtain dynamic area intrusion behavior data.

[0046] The present invention extracts climbing behavior, and the system can identify and capture the climbing action of individuals in the video, thereby effectively distinguishing normal behavior from potential security threats. The forbidden area division of climbing behavior can accurately identify which behaviors belong to cross-border behavior, especially in sensitive areas or prohibited areas, and timely discover intrusion behavior. This division helps to provide accurate alarms for security and take quick response measures. Further, through the evaluation of intrusion behavior routes, the system can analyze and predict the individual's action path, identify its intrusion trend in advance, and enhance the timeliness and accuracy of monitoring and early warning. Based on this, the intrusion behavior destination analysis can provide the system with more dimensional information, clarify the target direction or final destination of the intruder, and help formulate effective response strategies. Combined with the data of the intrusion behavior destination and the climbing prohibited area, the dynamic area intrusion behavior analysis ensures that the monitoring system can comprehensively and accurately identify the nature and scope of the intrusion behavior, effectively improve the security prevention ability, and reduce potential safety hazards. Through this series of analyses, the system realizes in-depth analysis and accurate identification of intrusion behavior, and improves the efficiency of emergency response.

[0047] Preferably, step S3 comprises the following steps:

[0048] Step S31: Perform video surveillance human flow statistics according to the video frame data of the video dynamic area to obtain video surveillance human flow data;

[0049] Step S32: Calculate the complexity of video surveillance behavior based on the video surveillance human flow data and the dynamic area behavior intention data to obtain video surveillance behavior complexity data;

[0050] Step S33: Calculate video stream overload according to the video surveillance behavior complexity data and the video surveillance human flow data to obtain surveillance video stream overload data;

[0051] Step S34: Perform monitoring video resource consumption growth evaluation based on the video monitoring behavior complexity data and the monitoring video stream overload data to obtain monitoring video resource consumption growth data.

[0052] The present invention performs human flow statistics on video frame data in the dynamic area of ​​the video. The system can monitor the density of people in the area in real time and identify key areas with high human flow, thereby improving the accuracy and response efficiency of monitoring. The behavior complexity calculation is combined with the human flow data and the dynamic area behavior intention data to comprehensively evaluate the activity density and complexity in the area, help the system determine the processing difficulty of the monitoring task, and provide a basis for resource scheduling and priority setting. Further, based on the behavior complexity and human flow data to calculate the overload of the video stream, it is possible to timely identify the problem of excessive data flow, prevent monitoring loss or delay caused by insufficient bandwidth or processing power, and combine the video complexity data and traffic overload data to evaluate the growth of resource consumption. The system can predict the resource demand when the load increases in real time, help to reasonably allocate processing power, and avoid system overload or crash. This series of steps helps to optimize the use of monitoring resources, improve system stability, and ensure that efficient operation and real-time response can be maintained under high load.

[0053] Preferably, step S33 includes the following steps:

[0054] Step S331: performing statistics on the crowd density exceeding the standard on the video surveillance crowd flow data to obtain the video surveillance crowd density exceeding the standard data;

[0055] Step S332: extracting the spatiotemporal distribution characteristics of dynamic crowd flow changes based on the video surveillance behavior complexity data and the video surveillance crowd flow density exceeding standard data, and obtaining dynamic crowd flow change distribution characteristic data;

[0056] Step S333: Calculate the dynamic spatial change amplitude of the dynamic change spatiotemporal distribution feature data and the video surveillance behavior complexity data to obtain dynamic spatial change frequency data;

[0057] Step S334: performing information overload processing according to the dynamic space change frequency data and the video surveillance crowd density exceeding standard data to obtain video surveillance information overload data;

[0058] Step S335: Perform video stream overload calculation based on the video surveillance information overload data to obtain surveillance video stream overload data.

[0059] The present invention performs density overload statistics on video surveillance human flow data, and the system can quickly identify areas where human flow is too dense, timely discover crowded or potentially dangerous situations, and provide support for safety management. Combined with behavioral complexity data and human flow density overload data to extract dynamic human flow change spatiotemporal distribution characteristics, the system can analyze the flow trend and density change of personnel in a specific area, and provide a basis for emergency warning and resource scheduling. Further, through the calculation of dynamic spatial change amplitude, the volatility and change frequency of human flow in different areas can be evaluated, which helps to judge the volatility of regional load and ensure timely response to emergencies. Based on the change frequency and density data, information overload processing can prevent the monitoring system from overloading information processing in high-density or complex scenes, ensuring that the monitoring data can be efficiently and accurately transmitted and processed. Through video flow overload calculation, the system can identify the situation where the flow exceeds the processable range, and timely adjust resources and load balance to avoid monitoring failure or delay caused by excessive flow. This series of steps ensures that the system can operate efficiently in complex and high-density monitoring scenarios, and guarantees the accuracy and real-time nature of monitoring data.

[0060] Preferably, step S4 comprises the following steps:

[0061] Step S41: performing monitoring behavior processing bottleneck assessment according to monitoring video resource consumption growth data to obtain monitoring behavior processing bottleneck data;

[0062] Step S42: performing monitoring behavior processing capacity attenuation analysis on the monitoring behavior processing bottleneck data and the monitoring video memory resource consumption data to obtain monitoring behavior processing capacity attenuation data;

[0063] Step S43: evaluating abnormal event response delay according to the monitoring behavior processing capacity attenuation data to obtain abnormal event response delay data;

[0064] Step S44: adjusting the digital monitoring video device according to the monitoring behavior processing capability attenuation data and the abnormal event response delay data to obtain digital monitoring device adjustment data.

[0065] The present invention performs a processing bottleneck assessment on the monitoring video memory resource consumption data, and the system can identify the links with excessive resource consumption, timely discover potential performance bottlenecks, and avoid processing delays or crashes in the monitoring system under high load conditions. By combining the processing bottleneck data with the memory consumption data to perform behavioral processing capacity attenuation analysis, the system can accurately predict the decline in processing capacity under a specific load, and give early warning so as to take measures to avoid the system from being unable to process complex monitoring tasks due to insufficient resources. Based on the attenuation data, the evaluation of the slow response to abnormal events can effectively measure the response speed of the system to emergencies under high load conditions, thereby discovering whether the monitoring system has the risk of delayed response, ensuring that it can quickly respond to security events, and adjusting the digital monitoring device through the attenuation data and the slow response data. The system can optimize the hardware and algorithm configuration according to the evaluation results, improve the processing capacity and resource utilization efficiency of the overall system, and ensure that it can still run efficiently and stably under high load conditions. This series of steps helps to improve the stability and emergency response capabilities of the monitoring system and ensure that resources are efficiently deployed at critical moments.

[0066] The present invention also provides a digital video device, comprising a digital video device main body, a power supply unit and an electrical control unit, wherein the power supply unit is installed inside the digital video device main body, the electrical control unit is electrically connected to the power supply unit, and the electrical control unit is used to charge the digital video device main body and control the digital video device main body, wherein the electrical control unit is used to execute the digital video device control method as described above, and the electrical control unit comprises:

[0067] Video dynamic area optimization module: obtain digital monitoring data; obtain real-time monitoring scenes based on digital monitoring data; obtain real-time monitoring scene data; optimize the dynamic area of ​​the monitoring video based on the real-time monitoring scene data to obtain the dynamic area optimization data of the monitoring video;

[0068] Regional behavior intention assessment module: read the dynamic region video frame according to the dynamic region optimization data of the surveillance video to obtain the video dynamic region video frame data; analyze the abnormal behavior of the dynamic region according to the video dynamic region video frame data to obtain the abnormal behavior data of the dynamic region; evaluate the behavior intention of the dynamic region according to the abnormal behavior data of the dynamic region to obtain the behavior intention data of the dynamic region;

[0069] Video resource consumption growth assessment module: calculates the complexity of video surveillance behavior based on video frame data in the video dynamic area and behavior intention data in the dynamic area to obtain video surveillance behavior complexity data; evaluates the growth of surveillance video resource consumption based on the video surveillance behavior complexity data to obtain surveillance video resource consumption growth data;

[0070] Digital surveillance video device adjustment module: perform surveillance behavior processing bottleneck assessment based on surveillance video resource consumption growth data to obtain surveillance behavior processing bottleneck data; perform surveillance behavior processing capacity attenuation analysis on surveillance behavior processing bottleneck data to obtain surveillance behavior processing capacity attenuation data; adjust the digital surveillance video device based on the surveillance behavior processing capacity attenuation data to obtain digital surveillance device adjustment data.

[0071] The present invention is to obtain digital monitoring data and perform real-time scene acquisition, which can ensure that the system can make timely responses based on the current monitoring environment and data dynamics. Through the capture and analysis of real-time scene data, the monitoring system can efficiently identify the areas that need to be focused on, and then optimize the dynamic areas of the monitoring video. This dynamic area optimization helps the system to concentrate resources to process data in key areas, without having to process indiscriminately in all areas, thereby improving processing efficiency and resource utilization. When further processing the video dynamic area frame data, abnormal behavior analysis can effectively detect potential abnormal or intrusion behaviors, ensuring that the system can make timely responses when emergencies occur. Through the subsequent analysis and evaluation of these abnormal behaviors, the system can not only identify the behavior pattern, but also infer the intention behind the behavior, which is of great value for the triggering of the early warning mechanism and the decision support of security. The evaluation of the dynamic area behavior intention enables the system to judge the threat level of the behavior, thereby dynamically adjusting the monitoring strategy and response priority. By combining the video dynamic area video frame data and the behavior intention data, the system can calculate the complexity of the video monitoring behavior, which is an important indicator for evaluating the monitoring data processing load. By calculating the complexity of monitoring behavior, the system can better understand the processing requirements of data, predict the processing pressure of the system, and avoid processing bottlenecks in high-complexity scenarios. As the complexity of behavior increases, the monitoring system will adjust according to the growth trend of resource consumption, and ensure that the system is always within the controllable range by timely evaluating the growth of resource consumption. Under high load conditions, the evaluation of the monitoring behavior processing bottleneck can identify the bottleneck of the system processing capacity and help system managers understand the performance problems encountered in specific situations. Next, the processing capacity decay analysis is performed to further reveal the performance decay phenomenon that occurs when the processing pressure is too large, and predict the system's extreme processing capacity. Through this step, the monitoring system can make timely adjustments when the load is too high, avoid system crashes or response delays, and ensure real-time performance and stability. According to the evaluation results of the monitoring behavior processing capacity decay, the system can adjust the digital monitoring video device, optimize the hardware and software configuration according to the actual resources and processing requirements, so as to ensure that it can continue to operate efficiently under high load conditions. Therefore, the present invention is an optimization process for a traditional digital video device control method. The traditional digital video device has the problem of inaccurate prediction of monitoring behavior and inaccurate prediction of monitoring behavior processing bottlenecks. The accuracy of monitoring behavior prediction and monitoring behavior processing bottleneck prediction has been improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 A schematic flow chart of the steps of a digital video device control method;

[0073] Figure 2 for Figure 1Detailed implementation steps of step S2 in the flowchart;

[0074] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0075] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0076] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0077] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0078] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0079] To achieve this, please refer to Figures 1 to 3 , a digital video device control method, comprising the following steps:

[0080] Step S1: acquiring digital monitoring data; acquiring real-time monitoring scenes according to the digital monitoring data; obtaining real-time monitoring scene data; optimizing the dynamic area of ​​the monitoring video according to the real-time monitoring scene data, and obtaining dynamic area optimization data of the monitoring video;

[0081] Step S2: Read the dynamic area video frame according to the dynamic area optimization data of the surveillance video to obtain the video dynamic area video frame data; perform dynamic area abnormal behavior analysis according to the video dynamic area video frame data to obtain the dynamic area abnormal behavior data; perform dynamic area behavior intention evaluation according to the dynamic area abnormal behavior data to obtain the dynamic area behavior intention data;

[0082] Step S3: Calculate the complexity of video monitoring behavior according to the video frame data of the dynamic area of ​​the video and the behavior intention data of the dynamic area to obtain the complexity data of video monitoring behavior; evaluate the growth of monitoring video resource consumption according to the complexity data of video monitoring behavior to obtain the growth data of monitoring video resource consumption;

[0083] Step S4: perform monitoring behavior processing bottleneck assessment based on monitoring video resource consumption growth data to obtain monitoring behavior processing bottleneck data; perform monitoring behavior processing capacity attenuation analysis on the monitoring behavior processing bottleneck data to obtain monitoring behavior processing capacity attenuation data; adjust the digital monitoring video device based on the monitoring behavior processing capacity attenuation data to obtain digital monitoring device adjustment data.

[0084] In the embodiment of the present invention, reference Figure 1 FIG. 1 is a flow chart of steps of a digital video device control method of the present invention. In this example, the digital video device control method includes the following steps:

[0085] Step S1: acquiring digital monitoring data; acquiring real-time monitoring scenes according to the digital monitoring data; obtaining real-time monitoring scene data; optimizing the dynamic area of ​​the monitoring video according to the real-time monitoring scene data, and obtaining dynamic area optimization data of the monitoring video;

[0086] In an embodiment of the present invention, digital monitoring data is obtained by multiple digital monitoring cameras installed in the target area. These cameras transmit the monitoring video stream to the data processing center by shooting the video data of the monitoring area in real time. The video data includes image frames, video streams and audio data. The monitoring camera uses a high-resolution video acquisition device, usually using a high-definition camera with a resolution of 1080p or higher for data acquisition to ensure the clarity of the image. While acquiring the data, the system will automatically perform a timestamp to ensure the time consistency and synchronization of the data. All acquired video data will first be pre-processed, including image denoising, color adjustment and image format conversion (such as RGB to YUV), and then stored in the central storage system, and real-time scene analysis will be performed based on the video data. The current scene is modeled using a background modeling algorithm in computer vision (such as a Gaussian mixture model) to identify the static background and dynamically changing areas in the scene. The identification basis of the dynamic area is the difference between the video frames (inter-frame difference). These areas are vehicles, pedestrians or dynamic objects in practical applications. By comparing the real-time monitoring scene with the background model, these dynamic areas can be extracted, and then the dynamic area optimization data of the monitoring video can be generated.

[0087] Step S2: Read the dynamic area video frame according to the dynamic area optimization data of the surveillance video to obtain the video dynamic area video frame data; perform dynamic area abnormal behavior analysis according to the video dynamic area video frame data to obtain the dynamic area abnormal behavior data; perform dynamic area behavior intention evaluation according to the dynamic area abnormal behavior data to obtain the dynamic area behavior intention data;

[0088] In the embodiment of the present invention, by processing the dynamic area optimization data extracted in step S1, the system accurately locates the dynamic area. The process uses an edge detection algorithm (such as the Canny operator) to extract boundaries, clarifies the contours and boundaries of dynamic objects through contour analysis, and determines their positions in the video frames. Video frames of the dynamic area will be read out from the storage, and these video frames contain all actions and events that occur in the area. Based on the read dynamic area video frames, the system further analyzes the video content through an action recognition algorithm (such as background subtraction) to detect abnormal behaviors in the area. Abnormal behaviors include but are not limited to running, fast movement, sudden changes, etc., which are different from normal scene activities. The key to behavior analysis is to determine whether these behaviors deviate from the preset normal mode, such as the difference between normal walking and sudden running behavior of pedestrians. After analysis, the system generates abnormal behavior data to indicate whether abnormal activities have occurred in the area. The system evaluates the behavior intention of the dynamic area by evaluating the contextual information of these abnormal behaviors. Based on the abnormal behavior data, simple rules are used to judge (such as walking behavior duration, speed, etc.) to evaluate the purpose and probability of the behavior. For example, fast running behavior indicates a safety risk or emergency situation. The system evaluates the intention of the behavior based on this data and generates dynamic regional behavior intention data.

[0089] Step S3: Calculate the complexity of video monitoring behavior according to the video frame data of the dynamic area of ​​the video and the behavior intention data of the dynamic area to obtain the complexity data of video monitoring behavior; evaluate the growth of monitoring video resource consumption according to the complexity data of video monitoring behavior to obtain the growth data of monitoring video resource consumption;

[0090] In an embodiment of the present invention, based on the dynamic area video frame data and dynamic area behavior intention data obtained in step S2, the system will calculate the behavior complexity of the entire monitoring area. The video monitoring behavior complexity calculation is based on the behavior pattern analysis, and the complexity of the behavior is comprehensively evaluated by detecting multiple dimensions such as the number of targets in the dynamic area, the target speed, and the interactive relationship between targets. The calculation method used in this step includes a target tracking algorithm (such as Kalman filtering, optical flow method) to monitor the movement trajectory of the target in real time and analyze whether there is mutual influence between the targets, such as intersection, collision or interaction.

[0091] Step S4: Based on the monitoring video resource consumption growth data, the monitoring behavior processing bottleneck is evaluated to obtain the monitoring behavior processing bottleneck data; the monitoring behavior processing capacity attenuation analysis is performed on the monitoring behavior processing bottleneck data to obtain the monitoring behavior processing capacity attenuation data; the digital monitoring video device is adjusted based on the monitoring behavior processing capacity attenuation data to obtain the digital monitoring device adjustment data. The system further combines the dynamic area behavior intention data for comprehensive calculation to evaluate the overall behavior complexity of the current monitoring area. For example, if there are multiple people moving quickly and interacting frequently with each other in the same dynamic area, the behavior complexity of the area is high. Through the complexity calculation, the system can determine the load of the current video monitoring environment, evaluate the processing requirements of video monitoring, and obtain the video monitoring behavior complexity data. Based on the above complexity data, the system will further evaluate the growth of monitoring video resource consumption. The evaluation of monitoring video resource consumption growth is mainly analyzed by the association between behavior complexity and processing requirements in the monitoring area. Complex behavior requires more computing resources, such as higher processor power, larger storage space and higher bandwidth. This evaluation process is calculated by the resource monitoring system (such as monitoring CPU, memory, bandwidth, etc.) to generate monitoring video resource consumption growth data.

[0092] Preferably, step S1 comprises the following steps:

[0093] Step S11: Acquire digital surveillance video data;

[0094] Step S12: acquiring the real-time scene of the monitoring video according to the digital monitoring video data; obtaining the real-time scene data of the monitoring video;

[0095] Step S13: collecting low-resolution monitoring area data according to the real-time scene data of the monitoring video to obtain low-resolution video monitoring area data;

[0096] Step S14: optimizing the dynamic region of the surveillance video according to the low-resolution region data of the video surveillance to obtain optimized dynamic region data of the surveillance video.

[0097] In an embodiment of the present invention, digital surveillance video data is collected in real time by multiple high-definition digital cameras installed in the surveillance area. These cameras are usually configured to support high-definition video streaming (such as 1080p or higher resolution). Each camera is installed in a suitable monitoring position and continuously captures all dynamic and static events in the surveillance area at a certain time interval (usually 30 frames per second). The video data is transmitted to the central processing unit via a wired or wireless network, using a standard video transmission protocol such as RTSP or RTMP to ensure data stability and low latency. The video stream includes not only image frame data, but also audio information (for example, a microphone that captures live sound). The video acquisition device has preset relevant parameters such as focal length, exposure time, white balance, etc. during installation to ensure that clear and distortion-free video data can be provided under different ambient lighting conditions. The digital surveillance video transmitted through the video data stream is subjected to real-time scene extraction. This process relies on image processing technology and uses computer vision algorithms to analyze each frame in the video stream. Through background modeling and foreground detection, the system identifies static background and dynamic foreground objects in the video. In this process, static areas are modeled using background modeling algorithms (such as Gaussian mixture models or ViBe algorithms), and changes in foreground objects are detected by differential methods. The goal of real-time scene acquisition is to extract meaningful dynamic event areas, such as moving pedestrians, vehicles, or moving objects, from the original video stream. This process includes graying the video frames, image differential processing, and then determining the position and shape of the dynamic area through connected region analysis to generate real-time scene data for surveillance videos. Each frame of the real-time scene data will be divided into several small areas, and different resolutions will be assigned to these areas according to the complexity and importance of the objects in the scene. Low-resolution areas are usually areas where there is no obvious dynamic change in the video frame or that do not contain key objects (such as background areas far away from the camera). The resolution of these areas is reduced through image area sampling and downsampling techniques (such as 2x2 block averaging or Gaussian pyramid method). This not only reduces the amount of data, but also effectively reduces the computational burden of subsequent processing. The downsampled low-resolution areas are stored and used as input data for subsequent dynamic area optimization. By analyzing the low-resolution area data, the system identifies the dynamic areas in the video. This process is implemented through edge detection algorithms (such as Canny edge detection or Sobel operator) to detect significant boundaries in the image and determine the outline of the object. The image in the low-resolution area is Gaussian blurred to reduce noise and smooth the image, and then edge detection is performed to extract important change information. By analyzing the low-resolution area data, the system identifies the dynamic areas in the video. This process is implemented through edge detection algorithms (such as Canny edge detection or Sobel operator) to detect significant boundaries in the image and determine the outline of the object.The image in the low-resolution area is Gaussian blurred to reduce noise and smooth the image, and then edge detection is performed to extract important change information. The system further optimizes the dynamic area based on the dynamic changes of the image and the object motion information. Through motion compensation and dynamic object tracking algorithms (such as Kalman filtering or particle filtering), the boundary detection of the dynamic area is optimized and false positives (i.e., misjudged dynamic areas) are removed. The system not only considers the changes in the static background, but also optimizes based on the motion pattern of the dynamic object.

[0098] Preferably, step S14 comprises the following steps:

[0099] Step S141: performing video grayscale image conversion according to the video surveillance low-resolution area data to obtain video area grayscale image data;

[0100] Step S142: performing dynamic region acquisition according to the video region grayscale image data to obtain video region grayscale image dynamic region data;

[0101] Step S143: performing dynamic region edge detection according to the video grayscale image dynamic region data to obtain dynamic region edge data;

[0102] Step S144: adjusting the edge detail resolution of the dynamic area edge data to obtain dynamic area edge detail adjusted data;

[0103] Step S145: performing foreground object separation according to the video grayscale image dynamic area data to obtain grayscale image foreground object separation data;

[0104] Step S146: Optimize the dynamic area of ​​the surveillance video based on the grayscale image foreground object separation data and the dynamic area edge detail adjustment data to obtain surveillance video dynamic area optimization data.

[0105] In an embodiment of the present invention, the low-resolution area data of the video surveillance is converted into a grayscale image. This process is achieved by converting the RGB value of each pixel of the color video frame into a single grayscale value. The specific operation is: for each pixel, its grayscale value is calculated, and the formula is Gray = 0.2989 × R + 0.5870 × G + 0.1140 × B, where R, G, and B represent the intensity values ​​of the three primary colors of red, green, and blue, respectively. Through this formula, the color information of the video frame is compressed into brightness information, the color dimension is removed, and the amount of calculation is reduced. The converted grayscale image retains the brightness and contrast information of the original video, while eliminating unnecessary color data, which is convenient for subsequent image analysis and processing. The obtained video area grayscale image data is used to collect dynamic areas based on the grayscale image data of the video area. By comparing the pixel value of each frame of the grayscale image with the pixel value of the previous frame, the significant change area in the image is identified. This process is achieved by image difference technology, that is, the grayscale image of the current frame is subtracted from the previous frame at the pixel level to calculate the difference between the two frames. Regions with large differences usually indicate dynamic objects or scene changes. Subsequently, morphological operations (such as corrosion and expansion) are used to eliminate noise and maintain the coherence of the changing regions. The system marks all regions with significant changes and uses them as dynamic regions to form the dynamic region data of the video grayscale image. The edge detection algorithm is used to analyze the dynamic region data of the video grayscale image to extract the significant boundaries in the image. The specific operation is to use the Sobel operator or the Canny edge detection algorithm to calculate the gradient change of each pixel in the image and detect the location where the pixel value changes sharply, so as to find the edge of the object. The Sobel operator calculates the gradient amplitude for each pixel by calculating the gradient in the horizontal and vertical directions, and determines the edge according to the threshold. The Canny algorithm includes multiple steps: Gaussian blurring the image to reduce noise; then calculating the gradient and extracting the edge; and then further refining the edge through non-maximum suppression and double threshold processing. Through these operations, the edges in the dynamic region are accurately identified and the dynamic region edge data is formed. The edges are fine-tuned using image scaling and interpolation algorithms. The edge area is enlarged to ensure that the details can be processed with higher resolution. Then, the edges of the image are smoothed by bilinear interpolation or cubic convolution interpolation algorithms to reduce jagged effects and improve edge clarity. This process can effectively enhance the performance of subtle changes in the image, especially in low-resolution areas. For cases where the edges are too blurred, the boundaries of the dynamic area can be more accurately identified after the detail resolution is adjusted. The adjusted dynamic area edge detail data is used to improve the accuracy of subsequent foreground object separation and the optimization effect of the dynamic area. Based on the difference between the significantly changing part of the dynamic area and the static background, the foreground object is extracted using background subtraction.Background subtraction estimates the background value of each pixel by building a background model (such as a Gaussian mixture model or a mean background model), and then compares each frame of the image with the background model to extract the foreground area. In this process, the extraction of foreground objects is based on the pixel areas in the dynamic area that are significantly different from the background. These areas usually represent dynamic events such as human movement and object movement. The foreground object separation data provides a basis for subsequent behavior analysis and event detection, and can accurately extract moving objects from the video. Use the foreground object separation data to determine the dynamic areas in the video, and combine these areas with the edge data after detail resolution adjustment to further improve the accuracy and detail performance of the dynamic area. In this process, by combining the position, size and edge information of the foreground object, the boundary of the dynamic area is optimized to match the actual position of the object more accurately. To achieve this goal, the system can use contour matching, morphological operations (such as dilation and erosion), and region growing algorithms to fine-tune the boundaries of the dynamic area.

[0106] Preferably, step S2 comprises the following steps:

[0107] Step S21: Read the dynamic area video frame according to the dynamic area optimization data of the monitoring video to obtain the video dynamic area video frame data;

[0108] Step S22: performing behavior recognition in the dynamic region on the video frame data of the video dynamic region to obtain behavior data of the video dynamic region;

[0109] Step S23: Performing dynamic area safety behavior analysis based on the video dynamic area behavior data to obtain dynamic area safety behavior data;

[0110] Step S24: performing dynamic area intrusion behavior analysis according to the video dynamic area behavior data to obtain dynamic area intrusion behavior data;

[0111] Step S25: performing dynamic area abnormal behavior analysis on the dynamic area intrusion behavior data and the dynamic area security behavior data to obtain dynamic area abnormal behavior data;

[0112] Step S26: Perform dynamic area behavior intention assessment based on the dynamic area abnormal behavior data to obtain dynamic area behavior intention data.

[0113] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0114] Step S21: Read the dynamic area video frame according to the dynamic area optimization data of the monitoring video to obtain the video dynamic area video frame data;

[0115] In an embodiment of the present invention, the dynamic area to be analyzed is extracted from the dynamic area optimization data of the monitoring video. The dynamic area is marked by reading the monitoring video stream frame by frame, and inter-frame processing is performed to determine the dynamic area in each frame. In specific implementation, the system accurately locates the position of the dynamic area by analyzing the timestamp of the video and the boundary information of the dynamic area, and then extracts the image frame data of the area. In this process, a video processing library (such as OpenCV) is used to read the video file or real-time video stream, and the key dynamic area in each frame is cut out according to the previous dynamic area optimization data.

[0116] Step S22: performing behavior recognition in the dynamic region on the video frame data of the video dynamic region to obtain behavior data of the video dynamic region;

[0117] In an embodiment of the present invention, the dynamic area in each frame is analyzed using action recognition technology. This process identifies the behavior pattern in the video by performing object detection and tracking on the dynamic area extracted from the frame. For example, motion analysis based on the optical flow method can be used to calculate the moving speed and direction of each pixel to detect the motion behavior of the target object. Then, by analyzing the movement direction, speed, and movement trajectory of the object, the behavior type present in the video, such as walking, running, jumping, etc., is identified. These behaviors are further confirmed by the classification algorithm, and dynamic area behavior data is generated.

[0118] Step S23: Performing dynamic area safety behavior analysis based on the video dynamic area behavior data to obtain dynamic area safety behavior data;

[0119] In the embodiment of the present invention, by analyzing the behavioral data in the dynamic area, it is determined whether there is normal or abnormal safety behavior. Specifically, the system classifies the behavior type of each dynamic area, such as walking, staying, turning around, etc. Then, the safety behavior rule library and the behavior model are used to compare these behaviors to determine whether the behavior meets the predetermined safety standards. For example, for work in the monitoring area, if the behavior meets the regulations (such as walking in a limited area and staying for no more than a certain time), it is marked as a safe behavior; if there is an abnormality in the behavior, such as wandering or staying in a certain location for a long time, it is marked as a potential dangerous behavior. After analysis, the system outputs the safety behavior data.

[0120] Step S24: performing dynamic area intrusion behavior analysis according to the video dynamic area behavior data to obtain dynamic area intrusion behavior data;

[0121] In an embodiment of the present invention, the behaviors in the dynamic area are classified in detail, and each behavior is evaluated by the set intrusion detection rules. For example, the system can monitor the behavior of entering or approaching these areas by setting virtual boundaries (such as fences or door areas). The system tracks the paths of objects in the dynamic area and detects whether there are objects crossing the set virtual boundaries, and then determines whether there is an intrusion. If abnormal behavior occurs in the dynamic area (such as crossing the security boundary or illegally entering a restricted area), the system will mark it as an intrusion and generate intrusion behavior data. This process predicts the motion trajectory of the object through a motion tracking algorithm (such as a Kalman filter).

[0122] Step S25: performing dynamic area abnormal behavior analysis on the dynamic area intrusion behavior data and the dynamic area security behavior data to obtain dynamic area abnormal behavior data;

[0123] In an embodiment of the present invention, the system combines the dynamic area intrusion behavior data with the dynamic area security behavior data to perform abnormal behavior analysis. By comparing the behavioral characteristics of the two, it is determined whether there is an abnormal situation. If the behavior in the dynamic area contains a mixed pattern of both intrusion behavior and security behavior, it is necessary to conduct an in-depth analysis of the behavior data. For example, if the intrusion behavior detected by the system occurs in an area with high traffic volume and is accompanied by behaviors such as staying or wandering for a long time, it can be determined as abnormal behavior, and the system will generate abnormal behavior data to indicate that there are safety hazards in the area. In addition, if the system detects a specific abnormal pattern when the intrusion behavior occurs (such as an object entering a restricted area and not leaving), it is considered to be a typical potentially dangerous behavior, and the monitoring and processing of abnormal behavior is further strengthened. This process can be combined with a rule engine and a machine learning algorithm to improve the accuracy of abnormal behavior recognition by accumulating historical data and comparing it with real-time data.

[0124] Step S26: Perform dynamic area behavior intention assessment based on the dynamic area abnormal behavior data to obtain dynamic area behavior intention data.

[0125] In an embodiment of the present invention, behavioral data of dynamic areas are collected and compared with historical data to analyze the potential intention of current behavior. For example, if an object is detected to repeatedly stay in a limited area and no predetermined normal behavior pattern occurs, it can be inferred that the behavioral intention of the object is to intend to carry out certain specific activities, such as destruction, theft, etc. In order to more accurately evaluate behavioral intentions, the system combines the previous security and intrusion analysis results to associate and infer the data of abnormal behaviors. If multiple behaviors occur at the same time, the system will use association rules or time series analysis methods to integrate the data of these behaviors, and predict their intentions based on historical patterns. The system generates dynamic area behavioral intention data.

[0126] Preferably, step S23 includes the following steps:

[0127] Step S231: collecting motion posture data according to the video dynamic area behavior data to obtain motion behavior posture data;

[0128] Step S232: Calculating the movement inclination according to the movement behavior posture data to obtain movement posture inclination data;

[0129] Step S233: performing a motion speed reduction calculation on the motion behavior posture data to obtain motion speed reduction data;

[0130] Step S234: performing a motion balance attenuation evaluation according to the motion action speed deceleration data and the motion posture inclination data to obtain motion balance attenuation data;

[0131] Step S235: Calculating the fall probability based on the sports balance attenuation data and the sports action speed reduction data to obtain sports fall probability data;

[0132] Step S236: Perform dynamic area safety behavior analysis based on the motion speed reduction data and the motion fall probability data to obtain dynamic area safety behavior data.

[0133] In the embodiment of the present invention, it is necessary to collect motion gestures based on the obtained video frame data of the dynamic area of ​​the video. The video frame data in the digital video device is used to identify the dynamic area in the video through image processing algorithms (such as background subtraction and contour extraction). Using this dynamic area information, a motion analysis technology based on the optical flow method is used for gesture detection. The optical flow method can infer the motion direction, speed and gesture changes of the object by analyzing the motion changes of pixels between consecutive video frames. The collected motion data will be converted into gesture data, including information such as the gesture angle and limb angle of the moving object, to form motion behavior gesture data. The key point positions of the target object are extracted, such as the head, shoulders, knees and other parts of a person. The inclination of the object or human body is calculated by the relative position relationship of these key points in three-dimensional space. In specific implementation, a geometric algorithm can be used to calculate the inclination of the object based on the angle between the key points. In particular, for the human body, the inclination is calculated by comparing the vertical angles between the main parts of the body (such as the angle between the upper body and the lower body) and the ground. This data generates motion posture inclination data, and the instantaneous speed of the moving object is calculated using the time interval between consecutive video frames and the displacement of the moving object in each frame. By analyzing multiple frames of data, the speed trend of the object is calculated, especially the deceleration trend. For the acceleration change of the object during the movement, the speed is analyzed using the acceleration calculation formula based on the differential method to obtain the degree of speed reduction. If the speed of the object slows down significantly during the movement, the movement slowdown can be inferred from the negative acceleration. This data is processed into motion movement speed reduction data. Based on the motion posture inclination data, it is determined whether the object or human body is in an unbalanced state. For example, when the tilt angle is large, the balance of the moving object will be affected. Secondly, according to the speed reduction data, if the speed of the object slows down quickly or the movement is incoherent, it indicates that its balance has further decreased. By combining these two data, the weighted average algorithm is used to perform attenuation evaluation on the balance of the object. The focus of the assessment is to determine whether there is a serious attenuation of balance due to slowed movement or excessive tilt angle, and output the motion balance attenuation data. Based on the balance attenuation data, if the balance of the moving object is significantly reduced, it indicates that the risk of the object (for example, the human body) losing balance during movement increases. Then, combined with the motion speed slowdown data, if the object's movement slows down faster (for example, a person's pace gradually slows down and approaches a stop), the risk of falling is further increased. Using a weighted comprehensive approach, the balance attenuation and speed slowdown data are combined, and the probability of falling is determined by setting a threshold, and then the fall probability data is calculated. Analyze the behavioral patterns reflected in the speed slowdown data. If there are large fluctuations in the speed slowdown and the probability of falling is high, it means that there is a safety risk in the movement of the object or human body.For the dynamic behavior of the human body, its safety status can be further evaluated by analyzing the changes in gait and the stability of the steps. If a clear deceleration trend and a high probability of falling are detected, it is identified as a potential dangerous behavior. Based on these data, safety behavior analysis is performed to determine whether there are abnormal safety behaviors in the area that require special attention. Dynamic regional safety behavior data is generated.

[0134] Preferably, step S24 comprises the following steps:

[0135] Step S241: extracting the climbing behavior according to the video dynamic area behavior data to obtain the video dynamic area climbing behavior data;

[0136] Step S242: dividing the video dynamic area climbing behavior data into climbing prohibited area behaviors to obtain climbing prohibited area behavior data;

[0137] Step S243: evaluating the intrusion behavior line according to the video dynamic area behavior data and the forbidden area crossing behavior data to obtain the intrusion behavior line data;

[0138] Step S244: performing intrusion behavior destination analysis based on the intrusion behavior route data to obtain intrusion behavior destination data;

[0139] Step S245: Perform dynamic area intrusion behavior analysis based on the intrusion behavior destination data and the prohibited area crossing behavior data to obtain dynamic area intrusion behavior data.

[0140] In the embodiment of the present invention, the motion object detection result in the video dynamic area behavior data is used to extract the action involving the climbing behavior. The video dynamic area behavior data has been identified by the motion detection algorithm (such as the optical flow method and the background subtraction method) to identify the dynamic activities in the area. Next, the behavior analysis algorithm is used, specifically based on morphological analysis and motion trajectory inference, to identify the action features related to the climbing behavior. For example, for the human body, by analyzing the limb motion trajectory, it is detected whether there is an action of climbing up from a lower object (such as a fence or a wall). If the moving object jumps, leaps or climbs quickly in the video frame, these features will be identified as climbing behavior. By using contour extraction and motion trend analysis, the time and dynamic area of ​​the climbing behavior are identified, the identification of the climbing behavior is formed, and the video dynamic area climbing behavior data is obtained. The area where climbing is prohibited is identified and the behavior is divided. The video dynamic area is divided into several sub-areas, including the allowed area and the prohibited area, using the existing monitoring system information and the preset geographic area model. The prohibited area can be defined by known physical obstacles (such as fences, walls or cordons) in the video surveillance image. Next, according to the area where the climbing behavior occurs, it is determined whether the behavior occurs in the prohibited area. If the climbing behavior occurs in the prohibited area, the system will mark this behavior as "violation" and classify it as climbing over the prohibited area. To ensure the recognition accuracy, the geometric analysis method can be used to accurately divide the area boundary according to the object's motion trajectory and climbing height, so as to avoid misjudgment. The association of all climbing behavior data with the prohibited area will form the climbing over the prohibited area behavior data. Evaluate whether there is an intrusion behavior. By analyzing the object trajectory and motion pattern in the dynamic area of ​​the video, combined with the boundary data of the prohibited area, the path analysis algorithm (for example, the path shortest distance algorithm based on dynamic programming) is used to evaluate the movement route of the intruder. If the climbing behavior occurs in the prohibited area, the system will identify the movement trajectory of the object and infer the route of its intrusion behavior. In specific implementation, the route of the intrusion behavior is determined by performing time series analysis on the continuous motion trajectory of the object in the video frame, that is, from the starting point of the object entering the prohibited area to the target area of ​​the invasion. In the evaluation process, the spatial position recognition technology can be combined to use coordinate transformation to compare the object's motion trajectory with the actual geographic coordinates to ensure the accuracy of the path evaluation. The intrusion behavior route data is formed, and the target area of ​​the intruder is determined by analyzing the intrusion behavior route, and the activity destination is predicted using spatial analysis technology. For example, the speed, direction and timestamp of the intruding object in the video frame are combined to infer where its destination is and whether it is a preset key area (such as a monitoring room, equipment room, etc.). The movement route is extended by the trajectory extension method to calculate the probability of the object reaching a certain target area. For the intruder's path, the regional heat map analysis method can also be used to evaluate the destination.This analysis can consider high-frequency areas of the intrusion path, past historical path data, etc. to help determine the destination of the intruder. The analysis results generate intrusion destination data, which is combined with the intruder's destination data to determine whether the intrusion constitutes a threat to a specific security area. If the intruder's destination is in a prohibited area or a critical security area, the system will mark it as a serious intrusion. On this basis, the system analyzes the specific pattern of the intrusion behavior and identifies the intruder's movement trajectory, activity mode, and target area. This process can be achieved through path prediction, dynamic monitoring area division, and danger level assessment. For example, if the intrusion occurs near an important facility, the system will further evaluate the risk level of the behavior and mark the intrusion behavior dynamically. In addition, the system will monitor video data in real time, combine historical behavior patterns, predict and identify behavior changes in dynamic areas, and obtain dynamic area intrusion behavior data.

[0141] Preferably, step S3 comprises the following steps:

[0142] Step S31: Perform video surveillance human flow statistics according to the video frame data of the video dynamic area to obtain video surveillance human flow data;

[0143] Step S32: Calculate the complexity of video surveillance behavior based on the video surveillance human flow data and the dynamic area behavior intention data to obtain video surveillance behavior complexity data;

[0144] Step S33: Calculate video stream overload according to the video surveillance behavior complexity data and the video surveillance human flow data to obtain surveillance video stream overload data;

[0145] Step S34: Perform monitoring video resource consumption growth evaluation based on the video monitoring behavior complexity data and the monitoring video stream overload data to obtain monitoring video resource consumption growth data.

[0146] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0147] Step S31: Perform video surveillance human flow statistics according to the video frame data of the video dynamic area to obtain video surveillance human flow data;

[0148] In an embodiment of the present invention, statistical information of human traffic is extracted from video frame data in a dynamic area of ​​a video. The specific implementation method is to identify the human targets appearing in the video frames through continuous image analysis of video frames, combined with background subtraction algorithms and target detection techniques (such as a classifier based on Haar features or a YOLO algorithm). The position of each human target in each frame after identification will be tracked, and the trajectory of each character will be determined by a tracking algorithm (such as a Kalman filter or a multi-target tracking algorithm). Subsequently, based on these trajectory data, the number of people entering the video surveillance area within a certain period of time is counted. To ensure accuracy, timestamps and inter-frame difference analysis are used to calculate the frequency of entry and exit of people in a specific area, and the number of people counted is organized in time series to form video surveillance human traffic data.

[0149] Step S32: Calculate the complexity of video surveillance behavior based on the video surveillance human flow data and the dynamic area behavior intention data to obtain video surveillance behavior complexity data;

[0150] In an embodiment of the present invention, the video surveillance human flow data obtained in step S31 is used in combination with the dynamic area behavior intention data to calculate the complexity of video surveillance behavior. By analyzing the time series of video surveillance human flow data, the density of human flow and the change trend in the monitoring area within a certain period of time are calculated. At the same time, the dynamic area behavior intention data provides the intention information of the behavior of people in the area, such as whether there is potential abnormal behavior, the density of activities, etc. By combining the human flow data with the behavior intention data, high-complexity areas can be identified, such as areas with dense personnel and accompanied by abnormal behavior. During the specific calculation, the weighted average method is used to combine the density change of human flow data and the behavior intention data to evaluate the regional behavior complexity. In high-density areas, the behavior complexity will increase with the increase in the number of people, and if accompanied by abnormal behavior (such as group gathering, fighting, etc.), the complexity will further increase. Finally, based on this calculation, the video surveillance behavior complexity data is obtained.

[0151] Step S33: Calculate video stream overload according to the video surveillance behavior complexity data and the video surveillance human flow data to obtain surveillance video stream overload data;

[0152] In an embodiment of the present invention, the resource requirements of the monitoring area are analyzed based on the video monitoring human flow data and behavior complexity data. Human flow data can reflect the density of people in the monitoring area, while behavior complexity data can reveal the types and frequencies of activities in the area. When the flow of people in the area increases, the monitoring video requires more bandwidth to handle more target identification, tracking and behavior analysis. In areas with higher complexity, such as when there are abnormal behaviors, frequently changing objects or fast-moving targets, the processing load of the video stream will also increase. Based on these data, the bandwidth consumption of the video stream is evaluated using a load prediction algorithm. When calculating the video stream overload, the behavior complexity and the human flow are weighted and summed, and then compared with the existing network bandwidth to determine whether bandwidth overload has occurred. The basis for determining overload is usually that the actual bandwidth demand of the video stream data exceeds the maximum bandwidth that the system can provide. Finally, monitoring video stream overload data is generated.

[0153] Step S34: Perform monitoring video resource consumption growth evaluation based on the video monitoring behavior complexity data and the monitoring video stream overload data to obtain monitoring video resource consumption growth data.

[0154] In an embodiment of the present invention, the monitoring video behavior complexity data reflects the number of targets in the video stream, the target movement speed, and the abnormal situation of the activity, which directly affects the encoding efficiency and processing load of the video stream. At the same time, the video stream overload data provides the actual bandwidth usage of the video stream under the current load. When evaluating the growth of resource consumption, the system will estimate the demand growth of video stream resources in the future period of time based on these data through a prediction model (such as a linear regression model or a timing analysis model). During the evaluation process, the current video resource consumption is calculated, and the growth rate of video resource consumption is predicted based on the changing trend of behavior complexity and the fluctuation of traffic flow. For example, if the density of people increases and the behavior complexity is high within a certain period of time, the system will predict that the resource consumption in the area will increase sharply. By comparing the remaining capacity of existing resources and the estimated resource demand, the monitoring video resource consumption growth data is finally generated.

[0155] Preferably, step S33 includes the following steps:

[0156] Step S331: performing statistics on the crowd density exceeding the standard on the video surveillance crowd flow data to obtain the video surveillance crowd density exceeding the standard data;

[0157] Step S332: extracting the spatiotemporal distribution characteristics of dynamic crowd flow changes based on the video surveillance behavior complexity data and the video surveillance crowd flow density exceeding standard data, and obtaining dynamic crowd flow change distribution characteristic data;

[0158] Step S333: Calculate the dynamic spatial change amplitude of the dynamic change spatiotemporal distribution feature data and the video surveillance behavior complexity data to obtain dynamic spatial change frequency data;

[0159] Step S334: performing information overload processing according to the dynamic space change frequency data and the video surveillance crowd density exceeding standard data to obtain video surveillance information overload data;

[0160] Step S335: Perform video stream overload calculation based on the video surveillance information overload data to obtain surveillance video stream overload data.

[0161] In an embodiment of the present invention, by analyzing the video surveillance human flow data, the human flow density in the video surveillance area in each time period is statistically calculated. The calculation of the human flow density is based on the number of targets in each frame image and combined with the actual size of the surveillance area. For example, if the area of ​​the surveillance area is A and the number of targets identified in a certain frame is N, the target density of the frame can be calculated to be N / A. Then, according to the preset threshold, it is determined whether the density exceeds the specified excess value. If the density exceeds the standard, it is considered that the human flow density exceeds the standard in the area. By comparing the density data of each time period, it is determined when and where the density exceeds the standard, and the density data of the time period and the corresponding area that exceed the standard are summarized to obtain the video surveillance human flow density excess data, and the video surveillance human flow density excess data obtained in step S331 is combined with the behavior complexity data obtained in step S32. By analyzing the spatiotemporal distribution of the human flow density excess data, the spatial concentration degree and the temporal fluctuation trend of the human flow are extracted. For example, if the density of a certain area exceeds the standard within a certain period of time and is accompanied by complex behaviors (such as pushing, running, etc.), it is considered that the change of the flow of people in this area has a high degree of complexity. In space, if the target density of a specific area exceeds the threshold and the complexity of the area is high, it means that the density of the flow of people in this area changes more drastically; in time, if the density change in this area has periodic fluctuations, it means that the time distribution characteristics of the flow of people are more obvious. Through spatiotemporal feature analysis, this information is sorted out to generate dynamic spatiotemporal distribution feature data of the flow of people. The spatiotemporal distribution feature data extracted in step S332 is combined with the behavior complexity data to determine the magnitude of the change in the density of people flow and the change in the complexity of behavior in a specific area within a certain period of time. For example, when the behavior complexity of a certain area is high, it is usually accompanied by dense crowd flow, and the spatial change magnitude is large at this time; conversely, if the behavior complexity is low, the spatial change magnitude is relatively small. The difference calculation method is used to obtain the amplitude of spatial changes in each time period, and the amplitude changes are weighted in combination with the complexity data to obtain the frequency data of dynamic spatial changes. The dynamic spatial change frequency data and the crowd density exceeding standard data are combined to calculate the overall information volume demand of the video stream in a certain area. If spatial changes occur frequently in a certain area and the crowd density in the area exceeds the standard, it means that the video stream in the area will generate a higher information volume demand. When processing information volume overload, the cumulative rate of information volume in each area in the exceeding standard state is calculated. Generally, the processing load of each frame can be obtained by calculating the relationship between the processing data volume of the video frame per unit time (including target detection, tracking and behavior recognition, etc.) and the regional density. When the information volume is overloaded, the processing capacity of the monitoring equipment will be limited, resulting in delays or data loss, and the video monitoring information volume overload data is obtained. The video monitoring information volume overload data is compared with the current bandwidth capacity of the system to calculate the difference between the bandwidth required for the video stream and the actual bandwidth of the system.If the required bandwidth exceeds the bandwidth that the current system can provide, it means that the video stream is overloaded. The specific calculation method is to estimate the bandwidth requirement of the video stream through the bandwidth calculation formula based on the resolution, frame rate, and target quantity and complexity of each area of ​​the video stream. If the bandwidth requirement exceeds the system bandwidth, the video stream will be overloaded. The system will mark the overloaded time period and area, and output the monitoring video stream overload data.

[0162] Preferably, step S4 comprises the following steps:

[0163] Step S41: performing monitoring behavior processing bottleneck assessment according to monitoring video resource consumption growth data to obtain monitoring behavior processing bottleneck data;

[0164] Step S42: performing monitoring behavior processing capacity attenuation analysis on the monitoring behavior processing bottleneck data and the monitoring video memory resource consumption data to obtain monitoring behavior processing capacity attenuation data;

[0165] Step S43: evaluating abnormal event response delay according to the monitoring behavior processing capacity attenuation data to obtain abnormal event response delay data;

[0166] Step S44: adjusting the device according to the monitoring behavior processing capability attenuation data and abnormal event response delay data to obtain digital monitoring device adjustment data.

[0167] In an embodiment of the present invention, the bottleneck evaluation of video processing is performed by monitoring the memory resource consumption data recorded by the monitoring system. The memory resource consumption data reflects the resource consumption of the video processing system in different time periods, including the storage of video frames, real-time decoding of video streams, calculation of target recognition and behavior analysis, etc. When the system resource consumption exceeds a preset threshold, the responsiveness of the video processing system is reduced, and even a processing bottleneck occurs. In the specific implementation process, a memory monitoring tool (such as monitoring software such as top or vmstat) is used to track the memory usage of the system in real time. Through the monitoring data, the amount of memory required for processing each frame of video is compared with the total available memory of the system. If the memory usage of the system in certain time periods reaches or approaches the maximum value, it is considered that there is a processing bottleneck in the time period. The bottleneck will cause delayed processing or frame loss of video data. Through a comprehensive analysis of memory resource consumption and system performance, the monitoring behavior processing bottleneck data is obtained, and the monitoring behavior processing bottleneck data and the memory resource consumption data are jointly analyzed. The processing bottleneck usually occurs when the system load is too high. When the video traffic increases or the complexity of the monitoring area increases, the consumption of memory resources will also increase. Using the relationship between memory usage and processing bottlenecks, the processing capacity changes of the calculation system at different memory consumption levels are calculated. For example, if the memory consumption exceeds 80%, the system processing capacity will decrease by 20%, and when the memory consumption is close to 100%, the processing capacity will decrease to less than 50%. Through this analysis method, the degree of attenuation of the system processing capacity can be quantified. Using the system's performance monitoring tools, such as perf (performance monitoring tool under Linux) or custom monitoring scripts, the system's processing capacity under high load conditions is measured and evaluated to obtain the monitoring behavior processing capacity attenuation data. The monitoring system should be able to respond to abnormal events in the video monitoring area in real time, such as intrusion, abnormal movement of objects, etc. If the system's processing capacity is attenuated, the response speed to abnormal events will be slowed down. During implementation, the delay of event response is calculated by real-time monitoring of the attenuation of the system's processing capacity and combining it with the specific time of occurrence of abnormal events. Specifically, if the system's behavior analysis module (such as motion detection, target recognition module) has a reduced processing capacity, the recognition and alarm of abnormal behaviors (such as climbing over, running, etc.) will be delayed. Use a delay monitoring tool (such as latencytop or a custom delay statistics tool) to record the time difference from the occurrence of each video event to the system processing and alarm. Through these data, the specific duration of the system's slow response is calculated, and abnormal event slow response data is generated. Based on the processing capacity decay data, evaluate whether the current system's hardware configuration can meet the video surveillance needs, especially the processing capacity under high load. For example, if the processing bottleneck occurs in the image processing, decoding or video storage module, you can consider upgrading the hardware of the corresponding module, or reduce the computing resource usage by optimizing the algorithm.Secondly, based on the abnormal event response delay data, evaluate the current system's response capability when abnormal events occur frequently. If the system's response speed drops significantly under high load, it is necessary to optimize the system's performance by increasing memory, improving processing power, or adjusting the video stream compression rate. Specific operations include adding more computing resources (such as CPU or GPU) or adjusting the video frame rate and resolution of the monitored area. By comprehensively considering the monitoring behavior processing capability and response delay data, the digital monitoring device adjustment data is finally output, including hardware configuration optimization suggestions, processing power improvement plans, etc.

[0168] The present invention also provides a digital video device, comprising a digital video device main body, a power supply unit and an electrical control unit, wherein the power supply unit is installed inside the digital video device main body, the electrical control unit is electrically connected to the power supply unit, and the electrical control unit is used to charge the digital video device main body and control the digital video device main body, wherein the electrical control unit is used to execute the digital video device control method as described above, and the electrical control unit comprises:

[0169] Video dynamic area optimization module: obtain digital monitoring data; obtain real-time monitoring scenes based on digital monitoring data; obtain real-time monitoring scene data; optimize the dynamic area of ​​the monitoring video based on the real-time monitoring scene data to obtain the dynamic area optimization data of the monitoring video;

[0170] Regional behavior intention assessment module: read the dynamic region video frame according to the dynamic region optimization data of the surveillance video to obtain the video dynamic region video frame data; analyze the abnormal behavior of the dynamic region according to the video dynamic region video frame data to obtain the abnormal behavior data of the dynamic region; evaluate the behavior intention of the dynamic region according to the abnormal behavior data of the dynamic region to obtain the behavior intention data of the dynamic region;

[0171] Video resource consumption growth assessment module: calculates the complexity of video surveillance behavior based on video frame data in the video dynamic area and behavior intention data in the dynamic area to obtain video surveillance behavior complexity data; evaluates the growth of surveillance video resource consumption based on the video surveillance behavior complexity data to obtain surveillance video resource consumption growth data;

[0172] Digital surveillance video device adjustment module: perform surveillance behavior processing bottleneck assessment based on surveillance video resource consumption growth data to obtain surveillance behavior processing bottleneck data; perform surveillance behavior processing capacity attenuation analysis on surveillance behavior processing bottleneck data to obtain surveillance behavior processing capacity attenuation data; adjust the digital surveillance video device based on the surveillance behavior processing capacity attenuation data to obtain digital surveillance device adjustment data.

[0173] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A digital video device control method, characterized in that: The following steps are involved: Step S1: Acquire digital monitoring data; acquire real-time monitoring scenes according to the digital monitoring data; Get real-time monitoring scene data; Optimize the dynamic area of ​​the surveillance video according to the real-time surveillance scene data to obtain the dynamic area optimization data of the surveillance video; Step S2: Read the dynamic area video frame according to the dynamic area optimization data of the monitoring video to obtain the video dynamic area video frame data; perform dynamic area abnormal behavior analysis according to the video dynamic area video frame data to obtain the dynamic area abnormal behavior data; Performing dynamic area behavior intention evaluation based on the dynamic area abnormal behavior data to obtain dynamic area behavior intention data; Step S3: Calculate the complexity of video surveillance behavior according to the video frame data of the video dynamic area and the behavior intention data of the dynamic area to obtain the complexity data of video surveillance behavior; According to the video surveillance behavior complexity data, the monitoring video resource consumption growth evaluation is performed to obtain the monitoring video resource consumption growth data; Step S4: performing monitoring behavior processing bottleneck assessment according to monitoring video resource consumption growth data to obtain monitoring behavior processing bottleneck data; Perform monitoring behavior processing capacity attenuation analysis on monitoring behavior processing bottleneck data to obtain monitoring behavior processing capacity attenuation data; The digital monitoring video device is adjusted according to the monitoring behavior processing capacity attenuation data to obtain digital monitoring device adjustment data.

2. The digital video device control method according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: Acquire digital surveillance video data; Step S12: acquiring the real-time scene of the monitoring video according to the digital monitoring video data; obtaining the real-time scene data of the monitoring video; Step S13: collecting low-resolution monitoring area data according to the real-time scene data of the monitoring video to obtain low-resolution video monitoring area data; Step S14: optimizing the dynamic region of the surveillance video according to the low-resolution region data of the video surveillance to obtain optimized dynamic region data of the surveillance video.

3. The digital video device control method according to claim 2, characterized in that: Step S14 includes the following steps: Step S141: performing video grayscale image conversion according to the video surveillance low-resolution area data to obtain video area grayscale image data; Step S142: performing dynamic region acquisition according to the video region grayscale image data to obtain video region grayscale image dynamic region data; Step S143: performing dynamic region edge detection according to the video grayscale image dynamic region data to obtain dynamic region edge data; Step S144: adjusting the edge detail resolution of the dynamic area edge data to obtain dynamic area edge detail adjusted data; Step S145: performing foreground object separation according to the video grayscale image dynamic area data to obtain grayscale image foreground object separation data; Step S146: Optimize the dynamic area of ​​the surveillance video based on the grayscale image foreground object separation data and the dynamic area edge detail adjustment data to obtain surveillance video dynamic area optimization data.

4. The digital video device control method according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: Read the dynamic area video frame according to the dynamic area optimization data of the monitoring video to obtain the video dynamic area video frame data; Step S22: performing behavior recognition in the dynamic region on the video frame data of the video dynamic region to obtain behavior data of the video dynamic region; Step S23: Performing dynamic area safety behavior analysis based on the video dynamic area behavior data to obtain dynamic area safety behavior data; Step S24: performing dynamic area intrusion behavior analysis according to the video dynamic area behavior data to obtain dynamic area intrusion behavior data; Step S25: performing dynamic area abnormal behavior analysis on the dynamic area intrusion behavior data and the dynamic area security behavior data to obtain dynamic area abnormal behavior data; Step S26: Perform dynamic area behavior intention assessment based on the dynamic area abnormal behavior data to obtain dynamic area behavior intention data.

5. The digital video device control method according to claim 4, characterized in that: Step S23 includes the following steps: Step S231: collecting motion posture data according to the video dynamic area behavior data to obtain motion behavior posture data; Step S232: Calculating the movement inclination according to the movement behavior posture data to obtain movement posture inclination data; Step S233: performing a motion speed reduction calculation on the motion behavior posture data to obtain motion speed reduction data; Step S234: performing a motion balance attenuation evaluation according to the motion action speed deceleration data and the motion posture inclination data to obtain motion balance attenuation data; Step S235: Calculating the fall probability based on the sports balance attenuation data and the sports action speed reduction data to obtain sports fall probability data; Step S236: Perform dynamic area safety behavior analysis based on the motion speed reduction data and the motion fall probability data to obtain dynamic area safety behavior data.

6. The digital video device control method according to claim 4, characterized in that: Step S24 includes the following steps: Step S241: extracting the climbing behavior according to the video dynamic area behavior data to obtain the video dynamic area climbing behavior data; Step S242: dividing the video dynamic area climbing behavior data into climbing prohibited area behaviors to obtain climbing prohibited area behavior data; Step S243: evaluating the intrusion behavior line according to the video dynamic area behavior data and the forbidden area crossing behavior data to obtain the intrusion behavior line data; Step S244: performing intrusion behavior destination analysis based on the intrusion behavior route data to obtain intrusion behavior destination data; Step S245: Perform dynamic area intrusion behavior analysis based on the intrusion behavior destination data and the prohibited area crossing behavior data to obtain dynamic area intrusion behavior data.

7. The digital video device control method according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: Perform video surveillance human flow statistics according to the video frame data of the video dynamic area to obtain video surveillance human flow data; Step S32: Calculate the complexity of video surveillance behavior based on the video surveillance human flow data and the dynamic area behavior intention data to obtain video surveillance behavior complexity data; Step S33: Calculate video stream overload according to the video surveillance behavior complexity data and the video surveillance human flow data to obtain surveillance video stream overload data; Step S34: Perform monitoring video resource consumption growth evaluation based on the video monitoring behavior complexity data and the monitoring video stream overload data to obtain monitoring video resource consumption growth data.

8. The digital video device control method according to claim 7, characterized in that: Step S33 includes the following steps: Step S331: performing statistics on the crowd density exceeding the standard on the video surveillance crowd flow data to obtain the video surveillance crowd density exceeding the standard data; Step S332: extracting the spatiotemporal distribution characteristics of dynamic crowd flow changes based on the video surveillance behavior complexity data and the video surveillance crowd flow density exceeding standard data, and obtaining dynamic crowd flow change distribution characteristic data; Step S333: Calculate the dynamic spatial change amplitude of the dynamic change spatiotemporal distribution feature data and the video surveillance behavior complexity data to obtain dynamic spatial change frequency data; Step S334: performing information overload processing according to the dynamic space change frequency data and the video surveillance crowd density exceeding standard data to obtain video surveillance information overload data; Step S335: Perform video stream overload calculation based on the video surveillance information overload data to obtain surveillance video stream overload data.

9. The digital video device control method according to claim 7, characterized in that: Step S4 includes the following steps: Step S41: performing monitoring behavior processing bottleneck assessment according to monitoring video resource consumption growth data to obtain monitoring behavior processing bottleneck data; Step S42: performing monitoring behavior processing capacity attenuation analysis on the monitoring behavior processing bottleneck data and the monitoring video memory resource consumption data to obtain monitoring behavior processing capacity attenuation data; Step S43: evaluating abnormal event response delay according to the monitoring behavior processing capacity attenuation data to obtain abnormal event response delay data; Step S44: adjusting the digital monitoring video device according to the monitoring behavior processing capability attenuation data and the abnormal event response delay data to obtain digital monitoring device adjustment data.

10. A digital video device, characterized in that: The digital video device comprises a main body, a power supply and an electrical control unit, wherein the power supply is installed inside the main body of the digital video device, the electrical control unit is electrically connected to the power supply, and the electrical control unit is used to charge the main body of the digital video device and control the main body of the digital video device, wherein the electrical control unit is used to execute the digital video device control method according to claim 1, and the electrical control unit comprises: Video dynamic area optimization module: obtain digital monitoring data; obtain real-time monitoring scenes based on digital monitoring data; obtain real-time monitoring scene data; optimize the dynamic area of ​​the monitoring video based on the real-time monitoring scene data to obtain the dynamic area optimization data of the monitoring video; Regional behavior intention assessment module: read the dynamic region video frame according to the dynamic region optimization data of the surveillance video to obtain the video dynamic region video frame data; analyze the abnormal behavior of the dynamic region according to the video dynamic region video frame data to obtain the abnormal behavior data of the dynamic region; evaluate the behavior intention of the dynamic region according to the abnormal behavior data of the dynamic region to obtain the behavior intention data of the dynamic region; Video resource consumption growth assessment module: calculates the complexity of video surveillance behavior based on video frame data in the video dynamic area and behavior intention data in the dynamic area to obtain video surveillance behavior complexity data; evaluates the growth of surveillance video resource consumption based on the video surveillance behavior complexity data to obtain surveillance video resource consumption growth data; Digital surveillance video device adjustment module: perform surveillance behavior processing bottleneck assessment based on surveillance video resource consumption growth data to obtain surveillance behavior processing bottleneck data; perform surveillance behavior processing capacity attenuation analysis on surveillance behavior processing bottleneck data to obtain surveillance behavior processing capacity attenuation data; adjust the digital surveillance video device based on the surveillance behavior processing capacity attenuation data to obtain digital surveillance device adjustment data.

Images