Automatic adjustment device for coverage range in distributed multi-camera environment

Abstract

The invention, which relates to the technical field of video monitoring and image data processing, provides a device for automatically adjusting a coverage area in a distributed multi-camera environment, comprising a camera main body, an image data processing terminal, a mounting top tray, a power-on motor base, a single camera controller, an adjusting holder and a camera. A data principal system is arranged in the image data processing terminal, the camera body comprises a mounting top tray, the bottom of the mounting top tray is in bolted connection with a power supply base, the interior of the power supply base is electrically connected with a circuit layer board, and the circuit layer board is electrically connected with a main power supply device and a standby power supply device. The invention provides a device for automatically adjusting a coverage area in a distributed multi-camera environment, and solves the problem that an existing monitoring system is insufficient in adaptive capacity in a dynamic environment. Through introduction of an advanced image data processing technology and fusion of various sensors, the system can analyze a personnel flow trend and environment change in a monitoring area in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of video monitoring and image data processing, and in particular to a device for automatically adjusting coverage in a distributed multi-camera environment. Background Art

[0002] In recent years, with the rapid development of intelligent monitoring technology, multi-camera monitoring systems have been widely used in public safety, traffic management, and environmental monitoring. These systems can achieve real-time monitoring of large areas through distributed multi-camera layouts, improving safety and response speed. Existing monitoring systems usually rely on fixed camera positions and angles and cannot be dynamically adjusted according to environmental changes, resulting in ineffective coverage of important areas or blind spots in some cases. In addition, traditional systems lack the ability to analyze personnel flow and environmental changes in real time, often rely on manual intervention, are inefficient, and cannot respond quickly to real-time situations.

[0003] Although some monitoring systems have introduced basic intelligent analysis modules, target detection and tracking functions, most systems still face the following shortcomings: First, they lack adaptive capabilities and cannot automatically adjust the coverage of the camera; second, the real-time data processing capabilities are insufficient, resulting in reduced recognition and tracking capabilities in high-density scenarios; third, there is a lack of effective coordination between data analysis modules, making it difficult to comprehensively consider environmental changes and target dynamics to achieve intelligent decision-making.

[0004] Therefore, a device for automatically adjusting coverage range in a distributed multi-camera environment is needed to solve the above problems. Summary of the invention

[0005] Technical issues solved In view of the shortcomings of the prior art, the present invention provides a device for automatically adjusting coverage in a distributed multi-camera environment, which solves the following problems: 1. Solved the problem of insufficient adaptability of existing monitoring systems in dynamic environments.

[0006] 2. It solves the problem of insufficient real-time data processing capabilities, which leads to reduced recognition and tracking capabilities in high-density scenarios.

[0007] 3. It solves the problem of lack of effective coordination between data analysis modules, making it difficult to comprehensively consider environmental changes and target dynamics and achieve intelligent decision-making.

[0008] Technical Solution To achieve the above objectives, the present invention is implemented through the following technical solutions: a device for automatically adjusting coverage in a distributed multi-camera environment, comprising a camera body, an image data processing terminal, a mounting top plate, a motorized base, a single camera controller, an adjustment pan and a camera, wherein the image data processing terminal is provided with a data main system, the camera body comprises a mounting top plate, the bottom of the mounting top plate is bolted to the motorized base, the motorized base is electrically connected to a circuit layer board inside, a main power supply device and a standby power supply device are electrically connected to the circuit layer board, the image data processing terminal, the mounting top plate, the motorized base, a single camera controller, an adjustment pan and a camera constitute the camera body, the top of the mounting top plate is connected to a support rod, the shape of the support rod is L-shaped, the bottom end of the support rod is connected to a power regulator, an energy storage device is installed at the bottom of the power regulator, one side of the energy storage device is electrically connected to a power transmission line, and one side of the power transmission line is electrically connected to a vibration energy collector.

[0009] Sp1, the vibration energy collector is a piezoelectric material, and the power regulator and energy storage device are combined to form a complete base shape of the support rod; Sp2, a power supply base is installed at the bottom of the mounting top plate, and the top of the power regulator is electrically connected to the standby power supply device in the power supply base through a line passing through the support rod; Sp3, both sides of the bottom of the power-on base and the front and back sides are electrically connected to a single camera controller, an adjustment platform is installed at the bottom of the single camera controller, and a camera is installed at the bottom of the adjustment platform; Sp4, the image data processing terminal receives and processes the image data from the camera, the power-on base is connected to the bottom of the mounting top plate and provides power support, the single camera controller controls the working status of each camera individually, and the adjusting pan-tilt adjusts the angle and direction of the camera.

[0010] Sp5. The data main system of the image data processing terminal includes an image receiving module, an image processing module, a target detection and tracking module, an environment change analysis module, a background modeling submodule, a coverage range analysis module, an adaptive adjustment module, and a data fusion and decision support module.

[0011] Preferably, the image receiving module receives real-time image data from four cameras, the image processing module processes the received image data, extracts feature information, and generates a visual monitoring area map, the target detection and tracking module is used to realize real-time target detection and multi-target tracking, the environmental change analysis module is used to analyze the personnel flow trend and environmental changes in the monitoring area, the background modeling submodule is used to dynamically analyze the scene background and eliminate interference factors, the coverage range analysis module analyzes the coverage range of the current camera and compares it with the preset optimal coverage range, the adaptive adjustment module dynamically optimizes the camera settings according to the monitoring data, and the data fusion and decision support module comprehensively analyzes multiple data sources and makes adjustment decisions.

[0012] Preferably, the target detection and tracking module is used to realize real-time target detection and multi-target tracking, and combines the YOLO algorithm to quickly identify dynamic objects in the monitoring area, predicts and tracks the target position through the Kalman filter algorithm, and performs multi-target tracking and data association through the DeepSORT algorithm.

[0013] Preferably, the environmental change analysis module is used to analyze the personnel flow trends and environmental changes in the monitoring area, and analyze the flow trends through a time series prediction algorithm, identify hot spots in the environment using a clustering algorithm, and detect and read abnormal data using an anomaly detection algorithm and trigger the camera to make adjustments.

[0014] Preferably, the background modeling submodule is used to dynamically analyze the scene background, eliminate interference factors, and improve the accuracy of target detection.

[0015] Preferably, the coverage range analysis module analyzes the coverage range of the current camera and compares it with a preset optimal coverage range to determine whether the angle or position of the camera needs to be adjusted.

[0016] Preferably, the adaptive adjustment module dynamically optimizes camera settings according to monitoring data, dynamically adjusts camera coverage according to fuzzy input through a fuzzy logic control algorithm, optimizes camera position and angle using a genetic algorithm, and learns the best adjustment strategy through a reinforcement learning algorithm.

[0017] Preferably, the data fusion and decision support module comprehensively analyzes multiple data sources and makes adjustment decisions, uses Bayesian network probabilistic reasoning and evaluates the success probability of the adjustment plan, uses a multi-sensor data fusion algorithm to comprehensively analyze the camera data and other sensor data, and finally models the camera through a decision tree algorithm to make adjustment decisions.

[0018] Preferably, the data management system has self-learning capabilities and continuously optimizes the algorithms and strategies of each module based on historical data and user feedback.

[0019] Preferably, the data management system controls and adjusts the pan / tilt to adjust the coverage of the camera according to the result of the system analysis.

[0020] Beneficial Effects

[0021] The present invention provides a device for automatically adjusting coverage in a distributed multi-camera environment. It has the following beneficial effects: 1. The present invention proposes an automatic adjustment device for coverage in a distributed multi-camera environment, which solves the problem of insufficient adaptability of existing monitoring systems in dynamic environments. By introducing advanced image data processing technology and integrating multiple sensors, the system can analyze the flow of people and environmental changes in the monitoring area in real time, and dynamically optimize the angle and position of the camera, thereby improving the comprehensiveness and accuracy of monitoring. The system can not only respond quickly to abnormal data reading, but also perform target detection and tracking through deep learning algorithms, YOLO, effectively improving the target recognition ability in complex scenes.

[0022] 2. Based on the design of the adaptive adjustment module and the data management system, the system has self-learning ability and can continuously optimize the algorithms and strategies of each module according to historical data and user feedback. This dynamic adjustment capability enables the system to always maintain the best coverage effect in different monitoring scenarios, significantly improving the intelligent level of security monitoring. At the same time, the system comprehensively analyzes multiple data sources through data fusion and decision support modules to provide monitoring personnel with accurate decision-making basis, so that the monitoring system has stronger response ability and flexibility when facing abnormal situations.

[0023] 3. The present invention proposes a more novel backup power supply design to ensure the stable operation of the camera. The system can better adapt to different environmental conditions. In construction sites or remote areas where power outages frequently occur, monitoring is always available, which improves safety and real-time performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a flow chart of the data processing system of the present invention; Figure 2 It is the overall structure diagram of the device of the present invention; Figure 3 This is a structural diagram of the camera body of the present invention; Figure 4 This is a diagram of the internal power supply structure of the power supply stand of the present invention; Figure 5 This is a design diagram of the internal circuit of the powered base of the present invention.

[0025] Legend: 1. Camera body; 2. Image data processing terminal; 3. Mounting top plate; 4. Power base; 5. Support rod; 6. Power regulator; 7. Energy storage device; 8. Transmission line; 9. Vibration energy collector; 10. Single camera controller; 11. Adjustment pan / tilt; 12. Camera; 13. Circuit board; 14. Main power supply device; 15. Standby power supply device. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention. Specific embodiment 1

[0028] Figure 1-5 As shown, an automatic adjustment device for coverage in a distributed multi-camera environment includes a camera body 1, an image data processing terminal 2, a mounting top plate 3, a motor base 4, a single camera controller 10, an adjustment pan head 11 and a camera 12. The image data processing terminal 2 is provided with a data main system. The camera body 1 includes a mounting top plate 3, the bottom of the mounting top plate 3 is bolted with the motor base 4, the inside of the motor base 4 is electrically connected with a circuit layer board 13, the circuit layer board 13 is electrically connected with a main power supply device 14 and a standby power supply device 15, the image data processing terminal 2, the mounting top plate 3, the motor base 4, the single camera controller 10, the adjustment pan head 11 and the camera 12 constitute the camera body 1, the top of the mounting top plate 3 is connected with a support rod 5, the shape of the support rod 5 is L-shaped, the bottom end of the support rod 5 is connected with a power regulator 6, the bottom of the power regulator 6 is installed with an energy storage device 7, one side of the energy storage device 7 is electrically connected with a transmission line 8, and a pile of the transmission line 8 is electrically connected with a vibration energy collector 9.

[0029] Sp1. The vibration energy collector 9 is made of piezoelectric material. The vibration energy collector 9 consists of the following main parts: first, the vibration sensing unit, which captures the mechanical vibration energy in the environment, including piezoelectric elements (piezoelectric ceramic sheets), electromagnetic induction units (using the relative movement of coils and magnets), and electrostatic sensors (based on capacitors).

[0030] The mechanical structure of the vibration energy harvester 9 converts environmental vibration into effective energy input through components such as springs, mass blocks, cantilever beams, and adjusts the resonant frequency of the system to adapt to the vibration source. Then the energy conversion module converts the captured mechanical energy into electrical energy, in which the piezoelectric material directly outputs electrical energy through the piezoelectric effect, and the electromagnetic device uses electromagnetic induction to generate current. Next is the power management circuit for rectification, boosting, voltage stabilization and energy storage, which saves the collected electrical energy through the energy storage unit (capacitor or micro battery) and provides a stable output. The output interface is responsible for transmitting the processed stable electrical energy to external devices or loads to meet actual application needs.

[0031] The power regulator 6 and the energy storage device 7 are combined to form a complete base shape of the support rod 5; Sp2, a power-on base 4 is installed at the bottom of the mounting top plate 3, and the top of the power regulator 6 is electrically connected to the standby power supply device 15 in the power-on base 4 through a line passing through the support rod 5; Sp3, both sides of the bottom of the power base 4 and the front and back are electrically connected to a single camera controller 10, an adjustment platform 11 is installed at the bottom of the single camera controller 10, and a camera 12 is installed at the bottom of the adjustment platform 11; Sp4, the image data processing terminal 2 receives and processes the image data from the camera, the power base 4 is connected to the bottom of the mounting top plate 3 and provides power support, the single camera controller 10 controls the working state of each camera 12 separately, and the pan / tilt 11 adjusts the angle and direction of the camera 12; Sp5, the data main system of the image data processing terminal 2 includes an image receiving module, an image processing module, a target detection and tracking module, an environment change analysis module, a background modeling submodule, a coverage range analysis module, an adaptive adjustment module, and a data fusion and decision support module.

[0032] The image receiving module receives real-time image data from four cameras 12, the image processing module processes the received image data, extracts feature information, and generates a visual monitoring area map, the target detection and tracking module is used to realize real-time target detection and multi-target tracking, the environmental change analysis module is used to analyze the personnel flow trend and environmental changes in the monitoring area, the background modeling submodule is used to dynamically analyze the scene background and eliminate interference factors, the coverage range analysis module analyzes the coverage range of the current camera and compares it with the preset optimal coverage range, the adaptive adjustment module dynamically optimizes the camera settings according to the monitoring data, and the data fusion and decision support module comprehensively analyzes multiple data sources and makes adjustment decisions.

[0033] The target detection and tracking module is used to realize real-time target detection and multi-target tracking, and combines the YOLO algorithm to quickly identify dynamic objects in the monitoring area, predicts and tracks the target position through the Kalman filter algorithm, and performs multi-target tracking and data association through the DeepSORT algorithm.

[0034] The environmental change analysis module is used to analyze the personnel flow trends and environmental changes in the monitoring area, and analyze the flow trends through the time series prediction algorithm, identify the hot spots in the environment using the clustering algorithm, and detect and read abnormal data using the anomaly detection algorithm and trigger the camera to make adjustments.

[0035] The background modeling submodule is used to dynamically analyze the scene background, eliminate interference factors, and improve the accuracy of target detection.

[0036] The coverage analysis module analyzes the current camera coverage and compares it with the preset optimal coverage to determine whether the camera angle or position needs to be adjusted.

[0037] The adaptive adjustment module dynamically optimizes the camera settings according to the monitoring data, dynamically adjusts the camera coverage according to the fuzzy input through the fuzzy logic control algorithm, optimizes the camera position and angle using the genetic algorithm, and learns the best adjustment strategy through the reinforcement learning algorithm.

[0038] The data fusion and decision support module comprehensively analyzes multiple data sources and makes adjustment decisions. It uses Bayesian network probabilistic reasoning and evaluates the success probability of the adjustment plan. It uses a multi-sensor data fusion algorithm to comprehensively analyze the camera data and other sensor data. Finally, it uses a decision tree algorithm to model and control the camera 12 to make adjustments.

[0039] The data management system has self-learning capabilities and continuously optimizes the algorithms and strategies of each module based on historical data and user feedback.

[0040] The data host system controls and adjusts the pan / tilt 11 to adjust the coverage of the camera 12 according to the result of the system analysis. Specific embodiment 2

[0042] Figure 1-5 As shown, in the present invention, the device itself has a main power supply and a backup power supply. In the application scenario of the present invention, the vibration energy collector 9 is buried in the ground. The working principle of the vibration energy collector 9 is to use transducer materials to convert the vibration energy in the environment into electrical energy. Piezoelectric materials are used here. The vibration energy collector 9 converts energy into electrical energy and then transmits it to the energy storage 7 through the transmission line 8. Here, the power regulator 6 and the energy storage 7 are combined to form the base of the support rod 5.

[0043] The circuit design of the entire monitoring device in the present invention includes the integration of multiple components. First, the camera 1 body is equipped with a power management circuit, which obtains power from the main power supply and the standby power supply, and distributes it to each camera and control module through a power regulator. In addition, the signal transmission circuit is responsible for connecting the camera 12 with the single camera controller 10 to ensure high-speed transmission of image data.

[0044] The image data processing terminal 2 integrates a high-performance microprocessor or FPGA to provide real-time image data processing capabilities. It is also equipped with a high-speed memory for temporary data storage and a storage interface for data backup. The input and output circuits provide an interface for connecting to external devices.

[0045] The connection circuit of the top plate 3 is installed to ensure that power is smoothly transmitted to the power base 4. The built-in power distribution circuit of the power base 4 distributes power to various components. It also has a status indication circuit to display the working status. The single camera controller 10 receives instructions from the image data processing terminal 2 through the control signal circuit and feeds back the status information of the camera 12.

[0046] The adjusting gimbal 11 controls the operation of the motor through the motor driving circuit, and cooperates with the sensor interface circuit to monitor the angle of the gimbal in real time. The power regulator 6 of the support rod 5 controls the power output of the energy storage device 7 through the power regulation circuit, and has a power monitoring function to ensure stable power supply.

[0047] The energy storage device 7 contains a charging management circuit for managing the energy collected from the vibration energy collector 9 to ensure safe charging. The vibration energy collector 9 is provided with an energy conversion circuit to convert the vibration energy into electrical energy to ensure stable operation of the device under different conditions.

[0048] Finally, the data fusion and decision support module integrates data from multiple sensors, processes it through sensor fusion circuits, and uses communication circuits to interact with external devices or cloud services to improve the intelligence level of the system.

[0049] Attached Figure 4 The circuit diagram of the main power supply device 14 and the standby power supply device 15 in the power-on base 4 shows the power conversion and management system of the monitoring device in the present invention, which is mainly used to convert the input 220V AC power into various low-voltage DC power required by the monitoring device. The circuit starts with 220V AC power and first passes through the primary and secondary EMI filters. These filters are used to eliminate electromagnetic interference and ensure the cleanliness and stability of the input power. Subsequently, the current enters the full-bridge rectifier filter module, which rectifies the AC power into DC power and outputs about +300V DC power.

[0050] The main power supply device in the circuit is responsible for converting the rectified high-voltage DC power into lower DC power to supply each subsystem. The conversion process relies on the main power switching transformer, which adjusts the input and output voltages through pulse width modulation (PWM). In this process, the driver transformer and driver tube are responsible for effectively transferring electrical energy to the transformer and ensuring efficient conversion of the system.

[0051] There is also a backup power system next to the main power supply, called the backup standby power supply device, which can take over the power supply when the main power fails. This part of the circuit provides +12V voltage through the standby power switching transformer, ensuring that the system can maintain basic functional operation even if the main power stops working.

[0052] The rectified and filtered power is further distributed to various voltage output terminals: +12V (yellow), +5V (white and red), -12V (blue), and +3.3V (orange). These voltage outputs meet the needs of different devices and provide multiple voltage support. The voltage stabilization and protection circuits in the figure are used to protect the system, ensure the stability of the output voltage, and provide overvoltage and overcurrent protection to prevent the system from being damaged by power fluctuations.

[0053] The entire circuit also contains multiple control signals. PG (Power Good) is used to indicate the power status, and the on / off control signal is used to turn the power on or off. The figure also mentions a +5VSB (Standby Voltage, purple), which is the standby voltage used to keep some circuits powered when the main system is turned off. Specific embodiment three

[0055] Figure 1-5 As shown, the key algorithms mentioned in Example 1 are analyzed in detail below: In the entire system, the YOLO algorithm is used to achieve real-time target detection and multi-target tracking. Its speed and efficiency are suitable for the dynamic needs of the monitoring system. The camera acquires images of the monitoring area in real time, and divides the images into grids through the YOLO model to detect and locate the target objects. The YOLO algorithm can directly output the category and location of the object (pedestrians, vehicles, etc.), and adjust the detection accuracy and speed of each target according to the system needs. The intersection-over-union (IOU) is used as an indicator to measure the overlap between the detection frame and the real frame, so that the module can maintain accurate detection in high-density scenes.

[0056] The formula of the YOLO algorithm is:

[0057] in is the grid size (a The grid is used to segment the image); The number of bounding boxes predicted for each mesh (usually 2 or 3), It is the intersection-over-union ratio, which is used to measure the overlap between the predicted box and the true box; and is the weight coefficient used to balance the coordinate loss and classification loss; is a true value, used to indicate whether the target exists in the grid; is the predicted probability, which indicates the possibility of the target category.

[0058] In the multi-target tracking module of the data processing system, the Kalman filter algorithm is used to predict and smooth the detected target position. The algorithm continuously updates the predicted position of the target based on the image frame sequence obtained by the camera, reduces noise interference and improves stability. The Kalman filter algorithm is based on prior state estimation and observation values, and can effectively process the movement trajectory of the target in a dynamic environment, and optimize the tracking results through the state transfer matrix and observation matrix. When the target is occluded, the algorithm can predict the possible position of the target, so that the system can continue to track after a short occlusion.

[0059] The formula is as follows: State transfer equation:

[0060] Observation equation:

[0061] in For the current moment The state vector of (the position and velocity of the target); The state transition matrix describes the system state from time to time Transfer to the moment ; is the control input matrix, describing the control input What affects the status? is the control input (acceleration and external force); is the process noise, is the current observation value, is the observation matrix that maps the state space to the observation space, is the observation noise.

[0062] In order to meet the multi-target tracking requirements that may exist in the system, the DeepSORT algorithm adds the Hungarian algorithm to the Kalman filter algorithm and realizes target association through distance measurement. This algorithm effectively solves the problems of multiple target overlap and occlusion in complex environments.

[0063] The formula is as follows:

[0064] DeepSORT combines detection and feature extraction modules to measure the Euclidean distance of target features, thereby matching and associating multiple targets. The camera inputs the features of multiple targets into the module, tracks the movement paths of different targets and ensures that there is no mistracking or missed tracking.

[0065] in For the goal and goals The distance measure between For the goal The center coordinates of For the goal The height and width of , For the goal The corresponding parameters.

[0066] The time series prediction algorithm (ARIMA or LSTM) in the monitoring system is used in the environmental change analysis module, which can predict the flow of people in the monitoring area based on historical data, so as to judge the possible events in the area in advance. The time series prediction algorithm can also detect the change trend.

[0067] The formula is as follows:

[0068] in Current moment Observed value of The constant term represents the base level; The autoregressive coefficient indicates the influence of the previous moment value on the current value; White noise (error term) at the current moment, wherein the error term includes white noise; The moving average coefficient represents the impact of the error at the previous moment on the current value.

[0069] The K-means clustering algorithm is used to identify hot spots in the monitoring area. In densely populated areas during peak hours, the system can analyze specific patterns in the area based on background modeling, automatically identify high-frequency areas, and cluster the location points in the monitoring data through the K-means algorithm to automatically identify hot spots. This algorithm can help the system dynamically adjust the camera's attention to hot spots and optimize the angle or focal length based on actual conditions.

[0070] The system divides the data into Classes make the samples of the same class closer and the distance between classes farther. The K-means algorithm formula is as follows:

[0071] in :No. Sample collection of classes :No. Center of the class : indicates sample The squared distance to the cluster center.

[0072] The anomaly detection algorithm is used in the environmental change analysis module to detect the status or abnormal activities in the area, such as sudden crowds or vehicle gatherings. Through methods such as Z-score, the system can detect phenomena where the behavior in the monitored area deviates from the normal range. The anomaly detection algorithm can set reasonable thresholds, such as the density or flow of people in the camera monitoring area. When the Z-score exceeds the preset value, the system will trigger an alarm signal and adjust the camera angle to more clearly observe the specific situation of the current status.

[0073] The Z-score method formula is as follows:

[0074] in The current observation value, The mean of the observations, Standard deviation, Z-score, fruit Usually considered abnormal.

[0075] The Bayesian network in the monitoring system plays a role in probabilistic reasoning in the decision support module. By combining multiple sensor data, the system can use Bayesian reasoning to calculate the probability of an event when it detects a possible abnormal situation, so as to help the system determine the response measures to be taken. After receiving the multi-sensor data of the current state, the system calculates the probability of the best response plan through Bayesian reasoning. When an abnormal crowd gathers in a certain area, the Bayesian network analyzes the previous data (time, historical traffic, etc.) to infer whether the nature of the event is a dangerous event or a temporary event, thereby providing the reliability of supporting decision-making.

[0076] The formula is as follows:

[0077] in :Given hour The posterior probability of :Given hour The likelihood probability of ; The prior probability of The marginal probability of .

[0078] The system uses a decision tree algorithm to make automatic decisions in the camera adjustment module. Based on the data collected by the camera and the environmental change information, the decision tree model can quickly determine whether the camera angle, position or zoom needs to be adjusted. The decision tree selects the best features based on information gain and models the adjustment factors (position, angle) of the camera in the system. If the hot spots in the monitoring area change frequently, the decision tree will make adjustment decisions based on factors such as the hot spot location, so that the camera coverage can adapt to the emergence of new hot spots.

[0079] The formula is as follows:

[0080] in is information gain, which is used to measure the selection of features The effect of splitting the dataset T; is the entropy of the data set, indicating uncertainty; Features All possible values ​​of ; For the characteristics The value is The data subset at that time; For subset The number of samples; For the dataset The total number of samples. Specific embodiment four: Figure 1-5 As shown, the following is a description of the specific application logic steps of each module and algorithm in the data management system of a distributed multi-camera environment coverage automatic adjustment device: The image receiving module consists of the following parts: Camera 12: a high-resolution surveillance camera with a wide-angle lens and support for real-time video transmission; Video capture card: converts the video signal of the camera into a digital signal, supports parallel input of multiple cameras, and ensures smooth transmission of data to subsequent processing modules; Network switch: used to connect multiple cameras to ensure efficient data transmission between different devices, support POE (Power over Ethernet) function, and simplify wiring.

[0082] The image processing module consists of the following parts: GPU (Graphics Processing Unit) is used to accelerate the calculation of deep learning algorithms such as YOLO and support high-concurrency image processing; Processor (CPU): Select Inteli7 high-performance multi-core processor to execute image processing and feature extraction algorithms to ensure real-time processing of data streams; Memory (RAM): High-speed memory that supports caching and fast access of large amounts of data, optimizing image processing speed.

[0083] The target detection and tracking module consists of the following parts: Deep learning accelerator card: NVIDIA Jetson series, specially designed for machine vision and deep learning, providing real-time target detection and tracking capabilities; Motion sensors: used to detect dynamic changes at the construction site and provide additional data support to improve the accuracy of target detection; Storage devices (SSD / HDD): High-speed storage devices are used to store real-time processed image data and historical records for subsequent analysis and playback.

[0084] The environmental change analysis module consists of the following components: Environmental sensors: temperature, humidity, air pressure and light sensors monitor construction environment changes in real time to ensure that the system can adjust monitoring strategies according to environmental conditions; Data processing unit: A processing unit that integrates sensor data and is composed of an Arduino microcontroller (composed of collecting and transmitting data to the main control system; Display: used to display environmental data and change trends in real time, which is convenient for on-site managers to monitor. The background modeling submodule consists of the following parts: Image storage device: high-capacity storage device (NAS) is used to store historical images for background modeling and analysis; Computing unit: A high-performance computer dedicated to running background modeling algorithms, with large memory and high-performance CPU to process large amounts of image data.

[0085] The coverage analysis module consists of the following components: Laser rangefinder: used to accurately measure the viewing angle and coverage of the camera to ensure the reasonable layout of the monitoring system; High-definition display: used to display the thermal map of camera coverage in real time, which is convenient for staff to make judgments and adjustments; Multi-function console: used to adjust camera position and settings, an operating interface that integrates multiple monitoring functions.

[0086] The adaptive adjustment module consists of the following parts: Servo motor: used to automatically adjust the angle and position of the camera to ensure the dynamic adaptability of the monitoring angle; Controller: integrated microprocessor, receives data in real time and controls the servo motor, and executes adaptive adjustment algorithm; Sensor network: monitors the status of the camera in real time and feeds back information to the control system to achieve closed-loop control.

[0087] The data fusion and decision support module consists of the following parts: High-performance server: used to process data from various modules, execute complex decision support algorithms, and has large memory and multi-core processing capabilities; Data acquisition module: used to collect data from different sensors and cameras, format the data uniformly and transmit it to the main server; Network interface: High-bandwidth network interface ensures smooth data transmission between modules and improves the overall efficiency of the system.

[0088] The data management system consists of the following parts: Central control computer: used to integrate and manage the operations of all modules, with high-performance CPU, sufficient memory and large storage space, supporting the running of complex self-learning algorithms.

[0089] Data backup equipment: used to regularly back up historical data to ensure data security and reliability and avoid data loss due to failures.

[0090] User interface device: Touch screen or tablet computer, providing a friendly user interface, allowing operators to monitor system status and adjust parameters in real time. Specific embodiment five

[0092] Figure 1-5 As shown, based on the above content, the following humanized designs can be carried out on the basis of the data management system: Fault detection and self-recovery mechanism: Adding a self-detection module to the equipment can monitor the status of each component (power supply, camera, sensor, etc.) in real time, and automatically switch to the backup system and alarm once a fault is detected.

[0093] Design self-recovery function to ensure automatic reinitialization after a failure.

[0094] Data security and privacy protection: Data encryption technology is introduced to ensure the security of transmitted image data and sensor data and prevent illegal access or tampering.

[0095] Develop privacy protection algorithms to blur non-target areas during monitoring and reduce infringement on personal privacy.

[0096] Energy Management System: Design an intelligent power management system to dynamically adjust power distribution based on the real-time operating status and environmental conditions of the equipment, ensuring efficient operation of each component and extending the service life of the equipment.

[0097] Introduce energy consumption monitoring function to provide real-time feedback of energy consumption data for subsequent optimization.

[0098] System Integration and Compatibility: Strengthen the modular design of the system to ensure the standardization of interfaces between modules, improve compatibility and scalability, and facilitate subsequent upgrades and maintenance.

[0099] Provides an API interface to allow third-party systems to integrate with the device and supports cross-platform data sharing.

[0100] User interface optimization: The intuitive and easy-to-use user interface provides real-time monitoring, data analysis and equipment management functions, ensuring operators can respond quickly.

[0101] Add visualization tools to help users understand monitoring data and system status and improve decision-making efficiency Experimentation and optimization: Conduct on-site testing and optimization, continuously adjust and improve system parameters based on feedback from the actual application environment, improve system performance and user experience, design test plans for a variety of application scenarios, and ensure the adaptability and reliability of the system in different environments.

[0102] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprising a reference structure" do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0103] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A device for automatically adjusting coverage in a distributed multi-camera environment, characterized in that: The invention comprises a camera body (1), an image data processing terminal (2), a mounting top plate (3), a power-on base (4), a single camera controller (10), an adjustment pan-tilt platform (11) and a camera (12); the image data processing terminal (2) is provided with a data main processing system; the camera body (1) comprises a mounting top plate (3); the bottom of the mounting top plate (3) is bolted to the power-on base (4); the power-on base (4) is electrically connected to a circuit board (13) inside the power-on base (4); the circuit board (13) is electrically connected to a main power supply device (14) and a standby power supply device (15); the image data processing terminal (2), the mounting top plate (3), the power-on base (4), the single camera controller (10), the adjustment pan-tilt platform (11) and the camera (12); 11) and a camera (12) constitute a camera body (1), the top of the mounting top plate (3) is connected to a support rod (5), the shape of the support rod (5) is L-shaped, the bottom end of the support rod (5) is connected to a power regulator (6), the bottom of the power regulator (6) is installed with an energy storage device (7), one side of the energy storage device (7) is electrically connected to a power transmission line (8), and the transmission line (8) is electrically connected to a vibration energy collector (9), and the data main system of the image data processing terminal (2) includes an image receiving module, an image processing module, a target detection and tracking module, an environmental change analysis module, a background modeling submodule, a coverage analysis module, an adaptive adjustment module, and a data fusion and decision support module.

2. According to the device for automatically adjusting coverage in a distributed multi-camera environment of claim 1, it is characterized in that: The image receiving module receives real-time image data from four cameras (12); the image processing module processes the received image data, extracts feature information, and generates a visual monitoring area map; the target detection and tracking module is used to realize real-time target detection and multi-target tracking; the environmental change analysis module is used to analyze the personnel flow trend and environmental changes in the monitoring area; the background modeling submodule is used to dynamically analyze the scene background and eliminate interference factors; the coverage range analysis module analyzes the coverage range of the current camera (12) and compares it with a preset optimal coverage range; the adaptive adjustment module dynamically optimizes the camera (12) settings according to the monitoring data; and the data fusion and decision support module comprehensively analyzes the data source and makes adjustment decisions.

3. According to the device for automatically adjusting coverage in a distributed multi-camera environment of claim 1, it is characterized by: The target detection and tracking module is used to realize real-time target detection and multi-target tracking, and combines the YOLO algorithm to quickly identify dynamic objects in the monitoring area, predict and track the target position through the Kalman filter algorithm, and perform multi-target tracking and data association through the DeepSORT algorithm; The formula of the YOLO algorithm is: ； in is the grid size; The number of bounding boxes predicted for each grid, It is the intersection-over-union ratio, which is used to measure the overlap between the predicted box and the true box; and is the weight coefficient used to balance the coordinate loss and classification loss; is a true value, used to indicate whether the target exists in the grid; is the predicted probability, which indicates the possibility of the target category; The Kalman filter algorithm formula is: State transfer equation: ； Observation equation: ； in For the current moment The state vector includes the position and velocity of the target; is the state transfer matrix, describing the system state from time Transfer to the moment ; is the input matrix, is the input variable The weight in the k-point system change; is the input variable, which is the external data input by the system at the k time position, and the external data includes the data of force, speed and direction; is the process noise, is the current observation value, is the observation matrix that maps the state space to the observation space, is the observation noise.

4. The device for automatically adjusting coverage in a distributed multi-camera environment according to claim 1, characterized in that: The environmental change analysis module is used to analyze the personnel flow trend and environmental changes in the monitoring area, and analyze the flow trend through a time series prediction algorithm, identify hot spots in the environment using a clustering algorithm, and detect and read abnormal data using an anomaly detection algorithm and trigger the camera (12) to make adjustments; The time series prediction algorithm formula is as follows: ； in Current moment Observed value of The constant term represents the base level; The autoregressive coefficient indicates the influence of the previous moment value on the current value; an error term at the current moment, wherein the error term includes white noise; The moving average coefficient represents the impact of the error at the previous moment on the current value.

5. The device for automatically adjusting coverage in a distributed multi-camera environment according to claim 1, characterized in that: The background modeling submodule is used to dynamically analyze the scene background and eliminate interference factors.

6. The device for automatically adjusting coverage in a distributed multi-camera environment according to claim 1, characterized in that: The coverage range analysis module analyses the coverage range of the current camera and compares it with a preset optimal coverage range to determine whether the angle and position of the camera (12) need to be adjusted.

7. The device for automatically adjusting coverage in a distributed multi-camera environment according to claim 1, characterized in that: The adaptive adjustment module dynamically optimizes the camera (12) settings according to the monitoring data, dynamically adjusts the camera (12) coverage according to the fuzzy input through a fuzzy logic control algorithm, optimizes the camera position and angle using a genetic algorithm, and learns the best adjustment strategy through a reinforcement learning algorithm.

8. The device for automatically adjusting coverage in a distributed multi-camera environment according to claim 1, characterized in that: The data fusion and decision support module comprehensively analyzes the data source and makes an adjustment decision, uses Bayesian network probability reasoning and evaluates the success probability of the adjustment plan, uses the sensor data fusion algorithm to comprehensively analyze the camera (12) data and other sensor data, and finally models the camera (12) adjustment decision through a decision tree algorithm.

9. The device for automatically adjusting coverage in a distributed multi-camera environment according to claim 1, characterized in that: The data management system has self-learning capabilities and optimizes the algorithms and strategies of each module based on historical data and user feedback.

10. The device for automatically adjusting coverage in a distributed multi-camera environment according to claim 1, characterized in that: The data main system controls and adjusts the pan / tilt platform (11) to adjust the direction of the camera (12) according to the result of the system analysis.

Images