A solar-powered intelligent video surveillance method and system

Abstract

The present invention discloses an intelligent video surveillance method and system based on solar-powered drive, which aims to improve the data transmission efficiency and power supply stability of solar-powered monitoring equipment. The method includes: constructing a data acquisition system for solar-powered monitoring equipment in the target area; acquiring the interaction status between the data acquisition system and the monitoring equipment in real time, and evaluating the data transmission quality based on the interaction status; optimizing the data transmission scheme according to the transmission quality, so as to efficiently obtain video monitoring screen data; analyzing the monitoring screen data, identifying abnormal events in the target area, and calculating the abnormal event probability value of each monitoring device; finally, according to the abnormal event probability value, dynamically adjusting the power supply strategy of the solar energy storage unit to ensure the efficient operation of the monitoring equipment under limited energy conditions. The present invention improves the reliability and sustainability of the monitoring system by intelligently scheduling energy supply, and is suitable for outdoor, remote areas and other scenarios with limited power supply.

Description

Technical Field

[0001] The present invention relates to the field of video surveillance technology, and in particular to an intelligent video surveillance method and system based on solar energy drive. Background Art

[0002] With the rapid development of artificial intelligence, the Internet of Things, and wireless communication technologies, intelligent video surveillance systems have gained widespread application in security, environmental monitoring, traffic management, and field research. Traditional video surveillance systems typically rely on wired power grids and fixed network transmission. However, in remote areas, uninhabited areas, and other special environments, power supply and network coverage are limited, making it difficult to meet the requirements for long-term stable operation. Therefore, solar-powered wireless intelligent video surveillance systems have become an important solution to this problem.

[0003] Existing solar-powered monitoring systems still face numerous challenges in data collection, transmission, and power management. On the one hand, due to the complex deployment environments of monitoring equipment, factors such as terrain and vegetation may affect the energy acquisition efficiency of solar panels, thereby affecting the system's ability to provide continuous power. On the other hand, monitoring equipment is widely distributed, and data transmission links may be affected by factors such as signal attenuation and obstruction, resulting in reduced data transmission quality. Furthermore, existing systems often adopt a fixed power supply strategy and fail to rationally allocate energy based on dynamic events in the monitoring area, resulting in low energy utilization efficiency and an inability to ensure the efficient operation of the monitoring system with limited solar resources.

[0004] In response to the above problems, there is an urgent need for an intelligent video surveillance method and system based on solar power drive. The invention should fully consider the impact of the geographical environment on solar power supply and data transmission, build an optimized data acquisition and transmission plan, and at the same time combine intelligent algorithms to dynamically adjust the energy supply strategy, improve the stability and energy efficiency of the monitoring system, and realize efficient video monitoring and abnormal event detection of the target area. Summary of the Invention

[0005] In order to solve at least one of the above technical problems, the present invention proposes an intelligent video surveillance method and system based on solar energy drive.

[0006] A first aspect of the present invention provides a solar-powered intelligent video surveillance method, comprising:

[0007] Build a data acquisition system for solar-powered monitoring equipment in the target area;

[0008] Acquire the interaction status between the data acquisition system and the monitoring device in real time, and determine the interaction quality between the data acquisition system and the monitoring device according to the interaction status;

[0009] Constructing a data transmission scheme between the data acquisition system and the monitoring equipment according to the interaction quality;

[0010] Acquire video surveillance data of solar-powered monitoring devices in a target area according to the data transmission scheme, identify abnormal events in the target area based on the video surveillance data, and output abnormal event probability values for each monitoring device;

[0011] The power supply strategy of the solar energy storage unit to each monitoring device is determined according to the abnormal event probability value.

[0012] In this solution, the data acquisition system for solar power supply monitoring equipment in the target area is constructed as follows:

[0013] Obtaining laser scanning data of a target area, the laser scanning data including terrain scanning data and vegetation scanning data, constructing a three-dimensional environmental simulation model of the target area based on the laser scanning data, obtaining installation location information of solar-powered monitoring devices in the target area, mapping each monitoring device to the three-dimensional environmental simulation model based on the installation location information, and constructing a monitoring environment simulation model of the target area;

[0014] Acquire historical data access data of each monitoring device in the target area, and determine data flow information of access to each monitoring device based on the historical data access data;

[0015] Normalize the installation location information and data flow information of each monitoring device to construct a normalized data input matrix;

[0016] A density clustering algorithm is introduced. The density clustering parameters, including the neighborhood radius ε and the minimum number of sample points MinPts, are set based on the data input matrix. Each data point in the data input matrix is initialized to a set of unvisited data points. A data point is randomly selected from the set of unvisited data points as the initial point. The point set in its ε-neighborhood is calculated. If the number of data points in the ε-neighborhood of the point is greater than or equal to MinPts, it is marked as a core point, and all points in its neighborhood are classified into the same cluster. Otherwise, they are marked as boundary points or noise points.

[0017] For each data point marked as a core point, all core points and density-reachable points in its ε-neighborhood are added to the current cluster until the cluster cannot be further expanded, so that all density-reachable points are completely classified into the corresponding cluster;

[0018] After all data points are accessed, clusters are output to obtain clustering results, the monitoring environment simulation model is divided into N three-dimensional grid units, and the data flow density of each three-dimensional network unit is determined according to the clustering results;

[0019] The installation location of the data server is determined according to the density of the data flow, and the data server is connected to each monitoring device to build a data acquisition system.

[0020] In this solution, the installation location of the data server is determined according to the density of the data flow, specifically:

[0021] Setting a data flow density threshold for a three-dimensional grid cell, traversing the data flow density data of each three-dimensional grid cell, and when the data flow density of a three-dimensional grid cell is greater than the density threshold, determining that the corresponding three-dimensional grid cell is a high data flow density area, and calculating the geometric center position of the high data flow density area as the installation location of the data server;

[0022] When the data flow density of the three-dimensional grid cell is not greater than the density threshold, obtaining data flow density data of adjacent three-dimensional grid cells; if the sum of the data flow density of the adjacent three-dimensional grid cells is greater than the density threshold, merging the adjacent three-dimensional grid cells into a data server deployment area, and calculating the geometric center position of the merged data server deployment area as the installation location;

[0023] Performing a pre-installation simulation on the installation location to obtain communication link signal strength data between the installation location and all monitoring devices in the target area. If the signal strength of any monitoring device is lower than a preset communication threshold, it is determined that an abnormal state exists between the monitoring device and the data server, preventing direct communication.

[0024] When a monitoring device that cannot transmit directly is detected, the terrain shielding parameters between the monitoring device and the installation location are extracted based on the three-dimensional environment simulation model, and the signal diffraction path loss value is calculated. If the path loss value exceeds the communication link tolerance threshold, the other devices in the cluster where the monitoring device is located are traversed to obtain the signal strength and geographic coordinate data of each device;

[0025] Calculate the straight-line distance between the monitoring device and other devices in the cluster based on the geographic coordinate data, select devices whose distance is less than the relay deployment radius as relay candidate nodes, and generate a relay priority score based on the remaining power and signal strength;

[0026] The device with the highest priority score is selected as the relay server, and the installation location of the data server and the location information of the relay server are output.

[0027] In this solution, the real-time acquisition of the interaction status between the data acquisition system and the monitoring device and the determination of the interaction quality between the data acquisition system and the monitoring device according to the interaction status are specifically as follows:

[0028] Real-time acquisition of communication link parameters between the data acquisition system and each monitoring device, including network delay value, signal strength fluctuation data and data packet loss rate, and constructing an interaction state matrix between the data acquisition system and the monitoring device based on the communication link parameters;

[0029] An interaction quality evaluation is performed on each monitoring device and the data acquisition system according to the interaction state matrix to obtain an interaction quality evaluation result.

[0030] In this solution, the data transmission solution between the data acquisition system and the monitoring device is constructed based on the interaction quality, specifically:

[0031] Obtain the data access status of the data acquisition system to each monitoring device, and obtain the interactive quality requirement data of the data acquisition system for data access to the monitoring device;

[0032] Determine the interaction quality redundancy index or interaction quality compensation index of each monitoring device according to the interaction quality requirement data, the interaction quality evaluation result, and the data access status, and construct an interaction quality redundancy-compensation index matrix using the interaction quality redundancy index or interaction quality compensation index of each monitoring device;

[0033] Determining a bandwidth allocation coefficient between each monitoring device and the data acquisition system according to the interaction quality redundancy-compensation index matrix;

[0034] The network transmission bandwidth allocation ratio between each monitoring device and the data acquisition system is determined according to the bandwidth allocation coefficient, and the network transmission bandwidth is allocated and transmitted to each monitoring device and the data interaction system according to the network transmission bandwidth allocation ratio to obtain a data transmission plan between the data acquisition system and the monitoring device.

[0035] In this solution, the video surveillance image data of the solar-powered monitoring device in the target area is obtained according to the data transmission solution, abnormal events in the target area are identified based on the video surveillance image data, and the abnormal event probability value of each monitoring device is output, specifically:

[0036] Based on the data transmission scheme, the video surveillance screen data uploaded by each monitoring device is received in real time, and the video surveillance screen data is decoded and preprocessed to extract dynamic target features and static background features in the video frame sequence;

[0037] Acquire historical dynamic target feature data of different abnormal event types, perform abnormal event type standard on the historical dynamic feature data, and obtain abnormal event labeled dynamic target feature data;

[0038] An abnormal event detection model is constructed based on a convolutional neural network, and a fully connected layer, an input layer, and a pooling layer of the abnormal event detection model are set. The mean square error is used as the loss function of the model, and the Adam optimization algorithm is used as the optimizer of the model. The abnormal event annotated target feature data is imported into the abnormal event detection model for learning and training;

[0039] The dynamic target features and static background features are imported into the learned and trained abnormal event detection model to perform static background area, and the dynamic target features are matched with the historical dynamic target feature data of different abnormal event types learned in the abnormal event detection model to determine the matching degree between the dynamic target features and the historical dynamic target feature data of different event types;

[0040] The abnormal event type of the image of each monitoring device is determined according to the matching degree, and the abnormal event probability value is output.

[0041] In this solution, the power supply strategy of the solar energy storage unit to each monitoring device is determined according to the abnormal event probability value, specifically:

[0042] Acquire the current remaining power data of the solar energy storage unit in real time, and divide the power supply priority of each monitoring device based on the probability value of the abnormal event;

[0043] When the probability value of the abnormal event is greater than the preset probability threshold, the corresponding monitoring device is marked as a high-priority device and is assigned a first power supply value; when the probability value of the abnormal event is not greater than the preset probability threshold, the corresponding monitoring device is marked as a normal-priority device and is assigned a second power supply value;

[0044] Calculate the sum of the first power supply values of all high-priority devices and the sum of the second power supply values of normal-priority devices to generate a total power supply requirement;

[0045] Determining whether the total power demand exceeds the maximum output power corresponding to the current remaining power data of the solar energy storage unit;

[0046] If it exceeds, the historical abnormal event probability values of all high-priority devices are traversed, the high-priority device with the lowest historical abnormal event probability value is downgraded to a regular priority device and the second power supply power value is reallocated, and iterative adjustment is made until the total power supply demand power is less than or equal to the maximum output power, and the power supply power of each monitoring device is output. The video surveillance screen acquisition resolution of each monitoring device is determined based on the power supply power, and the power supply strategy of the solar energy storage unit for each monitoring device is obtained.

[0047] A second aspect of the present invention further provides a solar-powered intelligent video surveillance system, comprising: a memory and a processor, wherein the memory includes a solar-powered intelligent video surveillance method program, and when the solar-powered intelligent video surveillance method program is executed by the processor, the following steps are implemented:

[0048] Build a data acquisition system for solar-powered monitoring equipment in the target area;

[0049] Acquire the interaction status between the data acquisition system and the monitoring device in real time, and determine the interaction quality between the data acquisition system and the monitoring device according to the interaction status;

[0050] Constructing a data transmission scheme between the data acquisition system and the monitoring equipment according to the interaction quality;

[0051] Acquire video surveillance data of solar-powered monitoring devices in a target area according to the data transmission scheme, identify abnormal events in the target area based on the video surveillance data, and output abnormal event probability values for each monitoring device;

[0052] The power supply strategy of the solar energy storage unit to each monitoring device is determined according to the abnormal event probability value.

[0053] The present invention discloses an intelligent video surveillance method and system based on solar-powered drive, which aims to improve the data transmission efficiency and power supply stability of solar-powered monitoring equipment. The method includes: constructing a data acquisition system for solar-powered monitoring equipment in the target area; acquiring the interaction status between the data acquisition system and the monitoring equipment in real time, and evaluating the data transmission quality based on the interaction status; optimizing the data transmission scheme according to the transmission quality, so as to efficiently obtain video monitoring screen data; analyzing the monitoring screen data, identifying abnormal events in the target area, and calculating the abnormal event probability value of each monitoring device; finally, according to the abnormal event probability value, dynamically adjusting the power supply strategy of the solar energy storage unit to ensure the efficient operation of the monitoring equipment under limited energy conditions. The present invention improves the reliability and sustainability of the monitoring system by intelligently scheduling energy supply, and is suitable for outdoor, remote areas and other scenarios with limited power supply. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 The present invention shows a flow chart of a solar-powered intelligent video surveillance method;

[0055] Figure 2 A flow chart showing the quality of interaction between a data acquisition system and a monitoring device according to the present invention is shown;

[0056] Figure 3 A flow chart showing a data transmission scheme constructed according to the present invention is shown;

[0057] Figure 4 The block diagram of the solar-powered intelligent video surveillance system of the present invention is shown. DETAILED DESCRIPTION

[0058] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0059] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0060] Figure 1 The flowchart of the solar-powered intelligent video surveillance method of the present invention is shown.

[0061] like Figure 1 As shown, the first aspect of the present invention provides an intelligent video surveillance method based on solar power drive, comprising:

[0062] S102, constructing a data acquisition system for solar power supply monitoring equipment in the target area;

[0063] S104, acquiring the interaction status between the data acquisition system and the monitoring device in real time, and determining the interaction quality between the data acquisition system and the monitoring device according to the interaction status;

[0064] S106, constructing a data transmission scheme between the data acquisition system and the monitoring device according to the interaction quality;

[0065] S108, acquiring video surveillance data of solar-powered monitoring devices in the target area according to the data transmission scheme, identifying abnormal events in the target area based on the video surveillance data, and outputting an abnormal event probability value for each monitoring device;

[0066] S110 , determining a power supply strategy of the solar energy storage unit to each monitoring device according to the abnormal event probability value.

[0067] It should be noted that by constructing a data acquisition system for solar-powered monitoring equipment in the target area, comprehensive collection and centralized management of the status and power supply conditions of each monitoring device can be achieved, ensuring that the system can continue to operate without the support of traditional power grids; the interaction status between the data acquisition system and the monitoring equipment is obtained in real time, and the interaction quality is evaluated based on this status, effectively reflecting the stability of the communication link and the operation status of the equipment; on this basis, a data transmission scheme is constructed to dynamically adjust the bandwidth allocation and transmission rate, thereby reducing data delay and improving the efficiency and reliability of video data transmission; further, the monitoring screen data is obtained in real time through this transmission scheme, and an intelligent algorithm is used to identify abnormal events in the target area, and the abnormal event probability value of each monitoring device is output, which significantly improves the accuracy and response speed of abnormal warning; finally, according to the abnormal event probability value, the power supply strategy of the solar energy storage unit for each monitoring device is reasonably determined, and priority power supply to key equipment and overall energy distribution optimization are achieved, thereby improving the overall energy efficiency and stability of the system, while reducing operating costs. The solar-powered monitoring device uses solar silicon wafers to effectively absorb solar radiation energy, converts light energy into electrical energy to provide power for the camera, has 4G full-network transmission, scheduled tasks, one-key watch, one-key cruise function, reads battery power information and performs OSD overlay.

[0068] According to an embodiment of the present invention, the data acquisition system for solar power supply monitoring equipment in the target area is constructed as follows:

[0069] Obtaining laser scanning data of a target area, the laser scanning data including terrain scanning data and vegetation scanning data, constructing a three-dimensional environmental simulation model of the target area based on the laser scanning data, obtaining installation location information of solar-powered monitoring devices in the target area, mapping each monitoring device to the three-dimensional environmental simulation model based on the installation location information, and constructing a monitoring environment simulation model of the target area;

[0070] Acquire historical data access data of each monitoring device in the target area, and determine data flow information of access to each monitoring device based on the historical data access data;

[0071] Normalize the installation location information and data flow information of each monitoring device to construct a normalized data input matrix;

[0072] A density clustering algorithm is introduced. The density clustering parameters, including the neighborhood radius ε and the minimum number of sample points MinPts, are set based on the data input matrix. Each data point in the data input matrix is initialized to a set of unvisited data points. A data point is randomly selected from the set of unvisited data points as the initial point. The point set in its ε-neighborhood is calculated. If the number of data points in the ε-neighborhood of the point is greater than or equal to MinPts, it is marked as a core point, and all points in its neighborhood are classified into the same cluster. Otherwise, they are marked as boundary points or noise points.

[0073] For each data point marked as a core point, all core points and density-reachable points in its ε-neighborhood are added to the current cluster until the cluster cannot be further expanded, so that all density-reachable points are completely classified into the corresponding cluster;

[0074] After all data points are accessed, clusters are output to obtain clustering results, the monitoring environment simulation model is divided into N three-dimensional grid units, and the data flow density of each three-dimensional network unit is determined according to the clustering results;

[0075] The installation location of the data server is determined according to the data flow density, and the data server is connected to each monitoring device to build a data acquisition system.

[0076] It should be noted that by introducing a density clustering algorithm to analyze the normalized monitoring equipment installation location information and historical data access flow data, the data flow density of each three-dimensional grid cell within the target area can be accurately divided, thereby determining the optimal installation location of the data server, achieving efficient data aggregation and transmission, and providing a basis for subsequent data scheduling and network optimization. Specifically, the laser scanning data, including terrain and vegetation scanning data, is used to construct a three-dimensional environmental simulation model of the target area, enabling a visual mapping of the spatial layout of the monitoring equipment. Normalization converts the equipment installation location and data flow information into a numerical matrix of uniform scale. The density clustering algorithm uses the neighborhood radius ε (which defines the influence range between data points) and the minimum number of sample points MinPts (the minimum number of points required to form a valid cluster) to classify data points, identifying dense areas and noise points. The data flow density reflects the distribution of data transmission load within each grid cell. Based on a set neighborhood radius ε and a minimum number of sample points, MinPts, the density clustering algorithm divides regions within the same local area where the number of data points reaches or exceeds a threshold into clusters. These clusters reflect the density of data flow transmission between monitoring devices within that area. By dividing the monitoring environment simulation model into multiple three-dimensional grid cells, the data flow density of each cell can be quantitatively described based on the number of clusters and data points contained within each grid. This not only accurately identifies areas with high data flow loads; the historical data access data includes data transmission volume, access frequency, time distribution, transmission delay, packet loss rate, etc.; the clustering result groups regions with similar data flow information within a preset range into one category.

[0077] According to an embodiment of the present invention, determining the installation location of the data server according to the data flow density is specifically:

[0078] Setting a data flow density threshold for a three-dimensional grid cell, traversing the data flow density data of each three-dimensional grid cell, and when the data flow density of a three-dimensional grid cell is greater than the density threshold, determining that the corresponding three-dimensional grid cell is a high data flow density area, and calculating the geometric center position of the high data flow density area as the installation location of the data server;

[0079] When the data flow density of the three-dimensional grid cell is not greater than the density threshold, obtaining data flow density data of adjacent three-dimensional grid cells; if the sum of the data flow density of the adjacent three-dimensional grid cells is greater than the density threshold, merging the adjacent three-dimensional grid cells into a data server deployment area, and calculating the geometric center position of the merged data server deployment area as the installation location;

[0080] It should be noted that by setting a data flow density threshold for three-dimensional grid cells and traversing the data flow density data of each grid cell, the area where the data flow density exceeds the threshold is automatically identified, and the geometric center of the area is used as the installation location of the data server, thereby ensuring that the data server can directly cover the high data flow density area; secondly, when the data flow density of a single grid cell does not reach the threshold, by integrating the data flow density data of adjacent grid cells, when the sum of adjacent cells exceeds the threshold, these cells are merged to form a deployment area, and the geometric center of the merged area is calculated as the installation location. This not only expands the coverage of the data server, but also ensures the integrity of the data flow density area and the efficiency of data transmission. The data server is a 4G wireless data receiving device.

[0081] Performing a pre-installation simulation on the installation location to obtain communication link signal strength data between the installation location and all monitoring devices in the target area. If the signal strength of any monitoring device is lower than a preset communication threshold, it is determined that an abnormal state exists between the monitoring device and the data server, preventing direct communication.

[0082] When a monitoring device that cannot transmit directly is detected, the terrain shielding parameters between the monitoring device and the installation location are extracted based on the three-dimensional environment simulation model, and the signal diffraction path loss value is calculated. If the path loss value exceeds the communication link tolerance threshold, the other devices in the cluster where the monitoring device is located are traversed to obtain the signal strength and geographic coordinate data of each device;

[0083] Calculate the straight-line distance between the monitoring device and other devices in the cluster based on the geographic coordinate data, select devices whose distance is less than the relay deployment radius as relay candidate nodes, and generate a relay priority score based on the remaining power and signal strength;

[0084] The device with the highest priority score is selected as the relay server, and the installation location of the data server and the location information of the relay server are output.

[0085] It should be noted that pre-installation simulation of the installation location allows for pre-deployment evaluation of the communication link status between the data server and all monitoring devices in the target area, ensuring that the pre-set installation location meets signal strength requirements during actual operation. If the signal strength of certain monitoring devices is found to be below a preset threshold, the system uses a three-dimensional environmental simulation model to extract terrain obstruction parameters between the monitoring device and the installation location and calculates the signal diffraction path loss value to determine whether the signal attenuation caused by terrain obstruction exceeds the communication link tolerance threshold. If excessive path loss is indeed present, the system traverses other devices within the cluster containing the monitoring device, obtains the signal strength and geographic coordinate data of each device, and uses this data to calculate the straight-line distance between the monitoring device and other devices, thereby selecting devices with distances less than the relay deployment radius as candidate relay nodes. Subsequently, a relay priority score is generated based on the remaining battery life and signal strength of the candidate devices. The device with the highest score is selected as the relay server, and the data server installation location and relay server location information are output. This process effectively overcomes direct transmission issues caused by terrain obstruction or distance limitations, ensuring stable and efficient data transmission in complex environments while optimizing network resource allocation and energy utilization.

[0086] Figure 2 A flow chart of determining the interaction quality between a data acquisition system and a monitoring device according to the present invention is shown.

[0087] According to an embodiment of the present invention, the real-time acquisition of the interaction status between the data acquisition system and the monitoring device, and the determination of the interaction quality between the data acquisition system and the monitoring device according to the interaction status, are specifically as follows:

[0088] S202, collecting communication link parameters between the data acquisition system and each monitoring device in real time, including network delay value, signal strength fluctuation data, and data packet loss rate, and constructing an interaction state matrix between the data acquisition system and the monitoring device based on the communication link parameters;

[0089] S204: Perform interaction quality evaluation on each monitoring device and the data acquisition system according to the interaction state matrix to obtain an interaction quality evaluation result.

[0090] It's important to note that this system collects communication link parameters such as network latency, signal strength fluctuations, and packet loss rates between the data acquisition system and each monitoring device in real time, organizing these key data by device or time series to form a multidimensional interaction status matrix. The system then performs a weighted analysis of each indicator in the matrix based on pre-set evaluation criteria and thresholds. For example, it compares network latency with tolerance values, signal fluctuations with stability indicators, and packet loss rates with allowable ranges. This comprehensive calculation yields an interaction quality score or grade for each monitoring device.

[0091] Figure 3 A flow chart of constructing a data transmission solution according to the present invention is shown.

[0092] According to an embodiment of the present invention, the data transmission scheme between the data acquisition system and the monitoring device constructed based on the interaction quality is specifically as follows:

[0093] S302, obtaining the data access status of the data acquisition system to each monitoring device, and obtaining the interaction quality requirement data of the data acquisition system for data access to the monitoring device;

[0094] S304: determining an interaction quality redundancy index or an interaction quality compensation index of each monitoring device according to the interaction quality requirement data, the interaction quality evaluation result, and the data access status, and constructing an interaction quality redundancy-compensation index matrix using the interaction quality redundancy index or the interaction quality compensation index of each monitoring device;

[0095] S306, determining a bandwidth allocation coefficient between each monitoring device and the data acquisition system according to the interaction quality redundancy-compensation index matrix;

[0096] S308, determining the network transmission bandwidth allocation ratio between each monitoring device and the data acquisition system according to the bandwidth allocation coefficient, performing network transmission bandwidth allocation transmission on each monitoring device and the data interaction system according to the network transmission bandwidth allocation ratio, and obtaining a data transmission plan between the data acquisition system and the monitoring device.

[0097] It should be noted that by comprehensively utilizing data access status, interaction quality demand data, and real-time interaction quality evaluation results, the interaction quality redundancy index or compensation index of each monitoring device is calculated, and a redundancy-compensation index matrix is constructed to obtain the bandwidth allocation coefficient between each device and the data acquisition system. This solution can dynamically adjust the network transmission bandwidth allocation ratio so that each monitoring device can meet the real-time data transmission requirements while optimizing the overall bandwidth utilization, avoiding network congestion or resource waste caused by data transmission differences between devices, and ensuring the stability and efficiency of video data transmission. The bandwidth allocation coefficient is positive when the interaction quality is redundant and negative when the interaction quality needs to be compensated. The larger the absolute value of the bandwidth allocation coefficient, the greater the interaction quality redundancy or the need for compensation. By allocating different network transmission bandwidths to each monitoring device, the unused network transmission bandwidth of the monitoring device with redundant interaction quality is reallocated to the monitoring device with compensated interaction quality, which can achieve dynamic balance and optimal utilization of network resources, thereby improving the stability and real-time performance of the overall data transmission. Specifically, this bandwidth reallocation mechanism provides additional support for devices in disadvantaged network conditions, reduces data transmission delays and packet loss rates, and enhances the robustness of monitoring data transmission links. At the same time, it also avoids the situation where bandwidth resources are excessive on some devices and insufficient on others, ensuring that the entire solar-powered intelligent video surveillance system can operate efficiently in various network environments and improving the accuracy and response speed of abnormal event detection.

[0098] According to an embodiment of the present invention, the video surveillance image data of the solar-powered monitoring device in the target area is obtained according to the data transmission scheme, abnormal events in the target area are identified based on the video surveillance image data, and the abnormal event probability value of each monitoring device is output, specifically:

[0099] Based on the data transmission scheme, the video surveillance screen data uploaded by each monitoring device is received in real time, and the video surveillance screen data is decoded and preprocessed to extract dynamic target features and static background features in the video frame sequence;

[0100] Acquire historical dynamic target feature data of different abnormal event types, perform abnormal event type standard on the historical dynamic feature data, and obtain abnormal event labeled dynamic target feature data;

[0101] An abnormal event detection model is constructed based on a convolutional neural network, and a fully connected layer, an input layer, and a pooling layer of the abnormal event detection model are set. The mean square error is used as the loss function of the model, and the Adam optimization algorithm is used as the optimizer of the model. The abnormal event annotated target feature data is imported into the abnormal event detection model for learning and training;

[0102] The dynamic target features and static background features are imported into the learned and trained abnormal event detection model to perform static background area, and the dynamic target features are matched with the historical dynamic target feature data of different abnormal event types learned in the abnormal event detection model to determine the matching degree between the dynamic target features and the historical dynamic target feature data of different event types;

[0103] The abnormal event type of the image of each monitoring device is determined according to the matching degree, and the abnormal event probability value is output.

[0104] It should be noted that by acquiring video surveillance footage in real time based on a data transmission solution and applying deep learning technology to detect abnormal events, this approach achieves accurate identification and efficient early warning of abnormal conditions in target areas. First, video data is decoded and preprocessed to extract dynamic target and static background features, thereby reducing environmental interference and improving target recognition accuracy. Second, by constructing an abnormal event detection model based on a convolutional neural network (CNN) and training it using the mean squared error (MSE) as a loss function and the Adam optimization algorithm, the model efficiently learns the characteristic data of different abnormal event types, improving recognition accuracy. Subsequently, by matching the dynamic target features in the current surveillance footage with the trained model and calculating the matching degree, the abnormal event type is accurately determined and the abnormal event probability value is output. This process not only enables rapid detection of sudden events but also classifies abnormal conditions of different levels based on probability values, helping the system prioritize high-risk events and improving the monitoring system's response speed and decision-making efficiency. Furthermore, this method adaptively learns and optimizes the anomaly detection model, continuously improving its recognition capabilities as more training data is collected. This maintains high detection accuracy even in complex environments, providing technical support for the efficient and stable operation of intelligent video surveillance systems.

[0105] According to an embodiment of the present invention, determining the power supply strategy of the solar energy storage unit to each monitoring device based on the abnormal event probability value is specifically as follows:

[0106] Acquire the current remaining power data of the solar energy storage unit in real time, and divide the power supply priority of each monitoring device based on the probability value of the abnormal event;

[0107] When the probability value of the abnormal event is greater than the preset probability threshold, the corresponding monitoring device is marked as a high-priority device and is assigned a first power supply value; when the probability value of the abnormal event is not greater than the preset probability threshold, the corresponding monitoring device is marked as a normal-priority device and is assigned a second power supply value;

[0108] Calculate the sum of the first power supply values of all high-priority devices and the sum of the second power supply values of normal-priority devices to generate a total power supply requirement;

[0109] Determining whether the total power demand exceeds the maximum output power corresponding to the current remaining power data of the solar energy storage unit;

[0110] If it exceeds, the historical abnormal event probability values of all high-priority devices are traversed, the high-priority device with the lowest historical abnormal event probability value is downgraded to a regular priority device and the second power supply power value is reallocated, and iterative adjustment is made until the total power supply demand power is less than or equal to the maximum output power, and the power supply power of each monitoring device is output. The video surveillance screen acquisition resolution of each monitoring device is determined based on the power supply power, and the power supply strategy of the solar energy storage unit for each monitoring device is obtained.

[0111] It should be noted that by acquiring the remaining power of the solar energy storage unit in real time and combining it with the abnormal event probability values of the monitoring equipment, power supply priorities are precisely divided, ensuring that stable power support is prioritized for high-risk areas under limited energy conditions, thereby improving the reliability and response speed of abnormal event monitoring. Secondly, by setting preset probability thresholds, monitoring devices are divided into high-priority and normal-priority groups, enabling hierarchical management of the power supply strategy and effectively preventing energy waste. Furthermore, when the total power demand exceeds the maximum output power of the current energy storage unit, priority is adjusted based on historical abnormal event probabilities, appropriately downgrading devices in low-risk areas, ensuring the rationality and sustainability of overall energy allocation. In addition, by adjusting the video surveillance image capture resolution of the monitoring equipment based on the power supply, high-quality video capture is prioritized in critical areas when energy is limited, while overall clarity is improved when energy is sufficient, thus balancing monitoring effectiveness and energy management. Ultimately, this method can adaptively adjust energy usage plans and improve the stability of the solar-powered intelligent video surveillance system.

[0112] According to an embodiment of the present invention, the further embodiment includes:

[0113] Real-time monitoring of the communication link parameters of each relay server in the target area, including signal strength and remaining power data. When it is detected that the signal strength of a relay server is lower than the preset communication threshold and the remaining power is lower than the power alarm threshold, the relay server is determined to have entered an abnormal state;

[0114] Extracting the geographic coordinates and remaining power data of all monitoring devices within the cluster where the abnormal relay server is located, screening devices with a remaining power higher than a preset relay activation threshold as candidate devices, and calculating the straight-line distance between each candidate device and the monitoring devices covered by the abnormal relay server;

[0115] Selecting a candidate device with the highest remaining battery and the shortest average distance to the covered device as a replacement relay server, sending a relay function activation instruction to the replacement relay server, and establishing a communication link between it and the data server and the original covered device;

[0116] Reduce the video transmission frame rate of the monitoring device whose abnormal event probability value in the cluster where the replacement relay server is located is lower than the preset threshold, mark the location of the abnormal relay server as a maintenance state, and update the device topology relationship table of the data acquisition system;

[0117] Continuously monitor the operating status of the alternative relay server until the abnormal relay server resumes power supply and the signal strength is stable, then switch the data transmission link back to the original relay server and release the relay function of the alternative relay server.

[0118] It should be noted that in solar video surveillance systems deployed in remote areas, when the relay server runs out of power due to equipment failure or extreme weather, the edge monitoring device responsible for forwarding will form a network island due to the interruption of the communication link. The traditional method that relies on manual maintenance has the problems of large response delay and high risk of data loss. The present invention monitors the signal strength and remaining power of the relay node in real time, and automatically selects candidate devices with sufficient remaining power and optimal geographical location from the cluster as alternative relays when an anomaly is detected, and quickly rebuilds the communication link; at the same time, it dynamically reduces the video transmission frame rate of low-risk monitoring devices to reduce the load pressure of the alternative relay, and seamlessly switches back to the original link after the original relay is restored. This solution realizes minute-level self-healing of relay failures, ensures network connectivity and continuous uploading of key monitoring data, significantly reduces monitoring blind spots and data loss rates caused by network islands, and at the same time extends the operating time of alternative relays through dynamic resource allocation, reducing the frequency of manual maintenance and system downtime.

[0119] Figure 4 The block diagram of the solar-powered intelligent video surveillance system of the present invention is shown.

[0120] A second aspect of the present invention further provides a solar-powered intelligent video surveillance system 4, comprising: a memory 41 and a processor 42, wherein the memory includes a solar-powered intelligent video surveillance method program, and when the solar-powered intelligent video surveillance method program is executed by the processor, the following steps are implemented:

[0121] Build a data acquisition system for solar-powered monitoring equipment in the target area;

[0122] Acquire the interaction status between the data acquisition system and the monitoring device in real time, and determine the interaction quality between the data acquisition system and the monitoring device according to the interaction status;

[0123] Constructing a data transmission scheme between the data acquisition system and the monitoring equipment according to the interaction quality;

[0124] Acquire video surveillance data of solar-powered monitoring devices in a target area according to the data transmission scheme, identify abnormal events in the target area based on the video surveillance data, and output abnormal event probability values for each monitoring device;

[0125] The power supply strategy of the solar energy storage unit to each monitoring device is determined according to the abnormal event probability value.

[0126] The present invention discloses an intelligent video surveillance method and system based on solar-powered drive, which aims to improve the data transmission efficiency and power supply stability of solar-powered monitoring equipment. The method includes: constructing a data acquisition system for solar-powered monitoring equipment in the target area; acquiring the interaction status between the data acquisition system and the monitoring equipment in real time, and evaluating the data transmission quality based on the interaction status; optimizing the data transmission scheme according to the transmission quality, so as to efficiently obtain video monitoring screen data; analyzing the monitoring screen data, identifying abnormal events in the target area, and calculating the abnormal event probability value of each monitoring device; finally, according to the abnormal event probability value, dynamically adjusting the power supply strategy of the solar energy storage unit to ensure the efficient operation of the monitoring equipment under limited energy conditions. The present invention improves the reliability and sustainability of the monitoring system by intelligently scheduling energy supply, and is suitable for outdoor, remote areas and other scenarios with limited power supply.

[0127] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0128] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed across multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0129] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0130] Those skilled in the art will appreciate that all or part of the steps of the above-mentioned method embodiments may be implemented by hardware associated with program instructions, and the aforementioned program may be stored in a computer-readable storage medium. When the program is executed, the program executes the steps of the above-mentioned method embodiments. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0131] Alternatively, if the integrated units described above are implemented as software modules and sold or used as standalone products, they can also be stored on a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product, stored on a storage medium, includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute all or part of the methods described in various embodiments of the present invention. The aforementioned storage media include various media capable of storing program code, such as removable storage devices, ROM, RAM, magnetic disks, or optical disks.

[0132] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A solar-powered intelligent video surveillance method, characterized in that: The following steps are involved: Build a data acquisition system for solar-powered monitoring equipment in the target area; Acquire the interaction status between the data acquisition system and the monitoring device in real time, and determine the interaction quality between the data acquisition system and the monitoring device according to the interaction status; Constructing a data transmission scheme between the data acquisition system and the monitoring equipment according to the interaction quality; Acquire video surveillance data of solar-powered monitoring devices in a target area according to the data transmission scheme, identify abnormal events in the target area based on the video surveillance data, and output abnormal event probability values for each monitoring device; Determining a power supply strategy for each monitoring device by the solar energy storage unit according to the abnormal event probability value; The data acquisition system for solar power supply monitoring equipment in the target area is constructed as follows: Obtaining laser scanning data of a target area, the laser scanning data including terrain scanning data and vegetation scanning data, constructing a three-dimensional environmental simulation model of the target area based on the laser scanning data, obtaining installation location information of solar-powered monitoring devices in the target area, mapping each monitoring device to the three-dimensional environmental simulation model based on the installation location information, and constructing a monitoring environment simulation model of the target area; Acquire historical data access data of each monitoring device in the target area, and determine data flow information of access to each monitoring device based on the historical data access data; Normalize the installation location information and data flow information of each monitoring device to construct a normalized data input matrix; A density clustering algorithm is introduced. The density clustering parameters, including the neighborhood radius ε and the minimum number of sample points MinPts, are set based on the data input matrix. Each data point in the data input matrix is initialized to a set of unvisited data points. A data point is randomly selected from the set of unvisited data points as the initial point. The point set in its ε-neighborhood is calculated. If the number of data points in the ε-neighborhood of the point is greater than or equal to MinPts, it is marked as a core point, and all points in its neighborhood are classified into the same cluster. Otherwise, they are marked as boundary points or noise points. For each data point marked as a core point, all core points and density-reachable points in its ε-neighborhood are added to the current cluster until the cluster cannot be further expanded, so that all density-reachable points are completely classified into the corresponding cluster; After all data points are accessed, clusters are output to obtain clustering results, the monitoring environment simulation model is divided into N three-dimensional grid units, and the data flow density of each three-dimensional grid unit is determined according to the clustering results; The installation location of the data server is determined according to the density of the data flow, and the data server is connected to each monitoring device to build a data acquisition system.

2. The solar-powered intelligent video surveillance method according to claim 1, characterized in that: The step of determining the installation location of the data server according to the density of the data flow is specifically as follows: Setting a data flow density threshold for a three-dimensional grid cell, traversing the data flow density data of each three-dimensional grid cell, and when the data flow density of a three-dimensional grid cell is greater than the density threshold, determining that the corresponding three-dimensional grid cell is a high data flow density area, and calculating the geometric center position of the high data flow density area as the installation location of the data server; When the data flow density of the three-dimensional grid cell is not greater than the density threshold, obtaining data flow density data of adjacent three-dimensional grid cells; if the sum of the data flow density of the adjacent three-dimensional grid cells is greater than the density threshold, merging the adjacent three-dimensional grid cells into a data server deployment area, and calculating the geometric center position of the merged data server deployment area as the installation location; Performing a pre-installation simulation on the installation location to obtain communication link signal strength data between the installation location and all monitoring devices in the target area. If the signal strength of any monitoring device is lower than a preset communication threshold, it is determined that an abnormal state exists between the monitoring device and the data server, preventing direct communication. When a monitoring device that cannot transmit directly is detected, the terrain shielding parameters between the monitoring device and the installation location are extracted based on the three-dimensional environment simulation model, and the signal diffraction path loss value is calculated. If the path loss value exceeds the communication link tolerance threshold, the other devices in the cluster where the monitoring device is located are traversed to obtain the signal strength and geographic coordinate data of each device; Calculate the straight-line distance between the monitoring device and other devices in the cluster based on the geographic coordinate data, select devices whose distance is less than the relay deployment radius as relay candidate nodes, and generate a relay priority score based on the remaining power and signal strength; The device with the highest priority score is selected as the relay server, and the installation location of the data server and the location information of the relay server are output.

3. The solar-powered intelligent video surveillance method according to claim 1, characterized in that: The real-time acquisition of the interaction status between the data acquisition system and the monitoring device, and the determination of the interaction quality between the data acquisition system and the monitoring device according to the interaction status, are specifically: Real-time acquisition of communication link parameters between the data acquisition system and each monitoring device, including network delay value, signal strength fluctuation data and data packet loss rate, and constructing an interaction state matrix between the data acquisition system and the monitoring device based on the communication link parameters; An interaction quality evaluation is performed on each monitoring device and the data acquisition system according to the interaction state matrix to obtain an interaction quality evaluation result.

4. The solar-powered intelligent video surveillance method according to claim 1, characterized in that: The data transmission scheme between the data acquisition system and the monitoring device is constructed according to the interaction quality, specifically: Obtain the data access status of the data acquisition system to each monitoring device, and obtain the interactive quality requirement data of the data acquisition system for data access to the monitoring device; Determine the interaction quality redundancy index or interaction quality compensation index of each monitoring device according to the interaction quality requirement data, the interaction quality evaluation result, and the data access status, and construct an interaction quality redundancy-compensation index matrix using the interaction quality redundancy index or interaction quality compensation index of each monitoring device; Determining a bandwidth allocation coefficient between each monitoring device and the data acquisition system according to the interaction quality redundancy-compensation index matrix; The network transmission bandwidth allocation ratio between each monitoring device and the data acquisition system is determined according to the bandwidth allocation coefficient, and the network transmission bandwidth is allocated and transmitted to each monitoring device and the data interaction system according to the network transmission bandwidth allocation ratio to obtain a data transmission plan between the data acquisition system and the monitoring device.

5. The solar-powered intelligent video surveillance method according to claim 1, characterized in that: The method further comprises: obtaining video surveillance image data of solar-powered monitoring devices in a target area according to the data transmission scheme, identifying abnormal events in the target area according to the video surveillance image data, and outputting an abnormal event probability value for each monitoring device, specifically: Based on the data transmission scheme, the video surveillance screen data uploaded by each monitoring device is received in real time, and the video surveillance screen data is decoded and preprocessed to extract dynamic target features and static background features in the video frame sequence; Acquire historical dynamic target feature data of different abnormal event types, perform abnormal event type standard on the historical dynamic target feature data, and obtain abnormal event labeled dynamic target feature data; An abnormal event detection model is constructed based on a convolutional neural network, and a fully connected layer, an input layer, and a pooling layer of the abnormal event detection model are set. The mean square error is used as the loss function of the model, and the Adam optimization algorithm is used as the optimizer of the model. The abnormal event annotated dynamic target feature data is imported into the abnormal event detection model for learning and training; Importing dynamic target features and static background features into the learned and trained abnormal event detection model to identify the static background area, and matching the dynamic target features with the historical dynamic target feature data of different abnormal event types learned in the abnormal event detection model to determine the matching degree between the dynamic target features and the historical dynamic target feature data of different event types; The abnormal event type of the image of each monitoring device is determined according to the matching degree, and the abnormal event probability value is output.

6. The solar-powered intelligent video surveillance method according to claim 1, characterized in that: The power supply strategy of the solar energy storage unit to each monitoring device is determined according to the abnormal event probability value, specifically: Acquire the current remaining power data of the solar energy storage unit in real time, and divide the power supply priority of each monitoring device based on the probability value of the abnormal event; When the probability value of the abnormal event is greater than the preset probability threshold, the corresponding monitoring device is marked as a high-priority device and is assigned a first power supply value; when the probability value of the abnormal event is not greater than the preset probability threshold, the corresponding monitoring device is marked as a normal-priority device and is assigned a second power supply value; Calculate the sum of the first power supply values of all high-priority devices and the sum of the second power supply values of normal-priority devices to generate a total power supply requirement; Determining whether the total power demand exceeds the maximum output power corresponding to the current remaining power data of the solar energy storage unit; If it exceeds, the historical abnormal event probability values of all high-priority devices are traversed, the high-priority device with the lowest historical abnormal event probability value is downgraded to a regular priority device and the second power supply power value is reallocated, and iterative adjustment is made until the total power supply demand power is less than or equal to the maximum output power, and the power supply power of each monitoring device is output. The video surveillance screen acquisition resolution of each monitoring device is determined based on the power supply power, and the power supply strategy of the solar energy storage unit for each monitoring device is obtained.

7. An intelligent video surveillance system based on solar energy drive, characterized in that: The solar-powered intelligent video surveillance system includes a storage device and a processor. The storage device includes a solar-powered intelligent video surveillance method program. When the solar-powered intelligent video surveillance method program is executed by the processor, the steps of a solar-powered intelligent video surveillance method as described in any one of claims 1 to 6 are implemented.

Images