Security monitoring method based on deployment and control ball data multiplexing

Abstract

The invention belongs to the technical field of intelligent monitoring, and provides a security monitoring method based on deployment and control ball data multiplexing, comprising the following steps: planning the monitoring range and number of deployment and control balls, and integrating Beidou positioning modules in the deployment and control balls; automatically adjusting the deploy and control ball to a horizontal state by using a holder and a tilt angle sensor; videos and data acquired by the monitoring balls are transmitted to a monitoring center in real time through wireless communication; preprocessing the transmitted video data, and performing intelligent analysis on the video data; videos and data are shared to related departments, multi-terminal access is allowed, and the dynamic state in a monitoring area is analyzed through a data multiplexing technology; when an abnormal condition is detected, an alarm mechanism is triggered immediately; according to the invention, not only are the coverage and real-time performance of monitoring improved, but also the efficiency and effectiveness of coping with security incidents are greatly improved through intelligent analysis and a data sharing mechanism, so that the reduction of potential safety hazards and risks, the optimization of resource configuration and the improvement of the overall security guarantee level are facilitated.

Description

Technical Field

[0001] The invention belongs to the technical field of intelligent monitoring, and in particular to a safety monitoring method based on control ball data multiplexing. Background Art

[0002] In today's society, with the continuous acceleration of urbanization and the frequent occurrence of various security incidents, security monitoring systems play a vital role in many fields such as public safety, traffic management, and commercial operations.

[0003] However, existing security monitoring systems usually rely on fixed cameras for video monitoring, which cannot flexibly adjust the viewing angle and respond to environmental changes in real time, and cannot effectively monitor dynamically changing scenes, resulting in limited monitoring effects;

[0004] Traditional monitoring systems rely on monitoring personnel to manually monitor multiple video signals, which is labor-intensive and easy to cause fatigue. For the processing of large amounts of monitoring data, monitoring personnel must not only keep an eye on each picture, but also judge and handle abnormal situations. In this case, any human negligence may lead to security incidents, resulting in serious consequences.

[0005] In existing monitoring systems, video data processing often lacks intelligence and mainly relies on manual playback and inspection of video recordings. This method is not only time-consuming and labor-intensive, but also difficult to achieve real-time anomaly detection and event response; and most monitoring solutions lack the ability to dynamically adjust monitoring strategies. They are highly dependent on pre-set monitoring logic and parameter configurations. When new risks or requirements arise due to environmental changes, monitoring strategies often cannot be adjusted in a timely manner, resulting in blind spots in security protection and poor response to emergencies.

[0006] To this end, technicians in this field have proposed a security monitoring method based on control ball data reuse, aiming to use intelligent analysis and data reuse technology to flexibly respond to dynamically changing environments, achieve accurate real-time monitoring and analysis, and enhance the overall dynamic response capability of monitoring and security prevention effects. Summary of the invention

[0007] In order to solve the above technical problems, the present invention provides a security monitoring method based on control ball data multiplexing to solve the problems raised in the technology.

[0008] A security monitoring method based on control ball data reuse includes the following steps:

[0009] S1. According to the monitoring requirements and spatial layout, plan the monitoring range and quantity of the control ball, integrate the Beidou positioning module into the control ball, and install the control ball at the predetermined location;

[0010] S2, using the pan / tilt and tilt sensor to automatically adjust the control ball to a horizontal state;

[0011] S3, transmit the video and data collected by the control ball to the monitoring center in real time through wireless communication;

[0012] S4, pre-processing the transmitted video data, performing intelligent analysis on the video data, and storing the processed video and data in a database;

[0013] S5. Share videos and data with relevant departments, allow multi-terminal access, and use data multiplexing technology to analyze dynamics within the monitoring area;

[0014] S6. When an abnormal situation is detected, the alarm mechanism is triggered immediately. The monitoring personnel will quickly initiate the emergency response plan based on the alarm information and conduct remote command and dispatch.

[0015] Preferably, the Beidou positioning module is used to record the position information of the control ball in real time, and use the pan / tilt head to perform horizontal and vertical rotation of the control ball, and combine with the inclination sensor to detect the inclination angle of the control ball, and automatically adjust to the optimal monitoring state.

[0016] Preferably, the transmitted video data is preprocessed, including denoising, image enhancement and frame extraction; deep features are extracted from the preprocessed image. Assuming that the input image is I, the convolution kernel is K, its size is F×F, and there are C′ filters, the convolution operation is expressed as:

[0017]

[0018] Among them, O(i,j) is the value of the output feature map at position (i,j), and (i+m,j+n) represents the movement of the convolution window on the input image;

[0019] After the convolution operation, the bias is added and processed by a nonlinear activation function, which is expressed as:

[0020] Z(i,j)=O(i,j)+b

[0021] A(i,j)=ReLU(Z(i,j))

[0022] Among them, Z(i,j) represents the value after adding the bias, b represents the bias term, and A(i,j) represents the output of the activation function;

[0023] Then perform a pooling operation to reduce the dimension of the feature map, expressed as:

[0024]

[0025] Among them, P(i,j) represents the value of the feature map at position (i,j) after pooling, S represents the size of the pooling window, and s is the stride.

[0026] Preferably, anomaly detection analysis is performed on the extracted features to identify abnormal behaviors, and normal feature representation is learned by reconstructing the input data. X is used as the extracted feature data set, then X = {x1, x2, ..., x n} as feature input, use the encoder to input feature x i Convert to implicit representation h i , and then the implicit representation h i Decode to reconstruct feature x i ′, calculate the reconstruction error for each sample, expressed as:

[0027]

[0028] Among them, d represents the dimension of the feature, and MSE represents the mean square error. According to the reconstruction error distribution of the training set, a threshold T is set. When the reconstruction error exceeds this threshold, it is marked as an abnormality. At the same time, a 3D convolutional network is used to directly process the video sequence, capture spatial and temporal features, and identify complex behavior patterns.

[0029] Preferably, the detected abnormal behaviors are classified, and the user behaviors are identified and labeled using the training model, including fighting, stealing, and large-scale gatherings; and the relationship between pedestrians, vehicles, and background in the scene is distinguished and understood through semantic segmentation;

[0030] The recognition results are combined with predefined security scenario rules to evaluate the recognized events. The rule engine makes judgments based on the rules. After identifying abnormal behaviors, the alarm mechanism is triggered, which automatically generates alarms and notifies monitoring personnel.

[0031] Preferably, data reuse technology is used to analyze the dynamics in the monitoring area, and long-term trend analysis and multi-dimensional evaluation are performed. Long-term trend analysis is to analyze the dynamically changing security risks by using historical and real-time data in the monitoring area; it analyzes historical data to identify long-term trends and patterns, and uses patterns in historical data to build a prediction model to identify security incidents that will occur in advance. The prediction model is expressed as: Where P(y=1X) represents the probability of event y under the condition of feature X, x1, x2, ..., x n represents the extracted features, β1,β2,...,β n is the model parameter.

[0032] Preferably, the multi-dimensional assessment provides a multi-dimensional perspective on security risks by integrating data from different monitoring points, time periods, and environmental conditions. After the results of the multi-dimensional analysis, a risk score is assigned to each area and time period, expressed as: Ri =w1·F event +w2·F env +w3·F time , where R i represents the risk score of the ith region, F event 、F env 、F time They represent event characteristics, environmental characteristics and time characteristics respectively. w1, w2 and w3 represent weight coefficients. High-risk areas and time periods are identified. According to real-time data and historical analysis results, security policies are dynamically adjusted to increase the monitoring frequency of high-risk areas.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] 1. After integrating the Beidou positioning module and the tilt sensor, the control ball can accurately locate and adjust the monitoring angle in real time to ensure the accuracy and timeliness of the monitoring data; through data multiplexing technology, the dynamics in the monitoring area can be analyzed to better identify potential threats, optimize resource allocation, and ensure dynamic monitoring of important moments and sites.

[0035] 2. The present invention can automatically identify abnormal behaviors by performing intelligent analysis and preprocessing on the transmitted video data, thereby reducing the workload of monitoring personnel and improving the speed of event response; when an abnormal situation is detected, the alarm mechanism can be quickly triggered so that the monitoring personnel can receive the alarm in time and take corresponding measures, thereby improving the emergency response capability.

[0036] 3. The present invention can reduce security risks through multi-dimensional analysis and intelligent monitoring methods, identify potential security risks through precise monitoring, and take measures early; and based on real-time data and historical analysis results, the monitoring strategy can be dynamically adjusted to respond to the ever-changing security threat environment in a timely manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The present invention is a flow chart of a security monitoring method based on control ball data reuse. DETAILED DESCRIPTION

[0038] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0039] As attached Figure 1 As shown:

[0040] Embodiment: The present invention provides a security monitoring method based on control ball data multiplexing, comprising the following steps:

[0041] S1. According to the monitoring requirements and spatial layout, plan the monitoring range and quantity of the control ball, integrate the Beidou positioning module in the control ball, and install the control ball at the predetermined location; the Beidou positioning module is used to record the location information of the control ball in real time.

[0042] S2. Use the pan / tilt head and tilt sensor to automatically adjust the control ball to a horizontal state; use the pan / tilt head to rotate the control ball horizontally and vertically, and use the tilt sensor to detect the tilt angle of the control ball to automatically adjust it to the best monitoring state.

[0043] S3. The video and data collected by the surveillance ball are transmitted to the monitoring center in real time through wireless communication. The real-time transmission of the video and data of the surveillance ball is realized through wireless communication, which reduces data delay and enables the monitoring center to quickly obtain on-site information.

[0044] S4, preprocess the transmitted video data, perform intelligent analysis on the video data, and store the processed video and data in the database; preprocess the transmitted video data, including denoising, image enhancement and frame extraction; extract deep features from the preprocessed image. Assuming that the input image is I, the convolution kernel is K, its size is F×F, and there are C′ filters, the convolution operation is expressed as:

[0045]

[0046] Among them, O(i,j) is the value of the output feature map at position (i,j), and (i+m,j+n) represents the movement of the convolution window on the input image;

[0047] After the convolution operation, the bias is added and processed by a nonlinear activation function, which is expressed as:

[0048] Z(i,j)=O(i,j)+b

[0049] A(i,j)=ReLU(Z(i,j))

[0050] Among them, Z(i,j) represents the value after adding the bias, b represents the bias term, and A(i,j) represents the output of the activation function;

[0051] Then perform a pooling operation to reduce the dimension of the feature map, expressed as:

[0052]

[0053] Among them, P(i,j) represents the value of the feature map at position (i,j) after pooling, S represents the size of the pooling window, and s is the stride. After multiple iterations of this process, the convolutional neural network automatically learns the deep features in the image, so that it can better represent the semantic information of the image and play an important role in subsequent classification and anomaly detection tasks.

[0054] Perform anomaly detection analysis on the extracted features to identify abnormal behaviors and learn normal feature representations by reconstructing the input data. X is used as the extracted feature data set, then X = {x1, x2, ..., x n} as feature input, use the encoder to input feature x i Convert to implicit representation h i , and then the implicit representation h i Decode to reconstruct feature x i ′, calculate the reconstruction error for each sample, expressed as:

[0055]

[0056] Among them, d represents the dimension of the feature, and MSE represents the mean square error. According to the reconstruction error distribution of the training set, a threshold T is set. When the reconstruction error exceeds this threshold, it is marked as abnormal. At the same time, a 3D convolutional network is used to directly process the video sequence to capture spatial and temporal features and identify complex behavior patterns. After deep learning feature extraction, the above unsupervised learning method can effectively detect abnormal behaviors and security threats, and the use of a 3D convolutional network can effectively capture spatial and temporal features in time series and extract complex behavior patterns.

[0057] Classify detected abnormal behaviors and use trained models to identify and annotate user behaviors, including fighting, stealing, and large-scale gatherings; and use semantic segmentation to distinguish and understand the relationship between pedestrians, vehicles, and background in the scene;

[0058] Combining the recognition results with the predefined security scenario rules, the recognized events are evaluated, and the rule engine makes judgments based on the rules. After identifying abnormal behaviors, the alarm mechanism is triggered, and an alarm is automatically generated and notified to the monitoring personnel. Step S4 focuses on the intelligent analysis of the video data itself, behavior recognition, and instant alarms in the monitoring scene. The main data source is real-time video streams, focusing on image processing and behavior recognition.

[0059] S5. Share videos and data with relevant departments, allow multi-terminal access, and use data reuse technology to analyze the dynamics in the monitoring area; use data reuse technology to analyze the dynamics in the monitoring area, conduct long-term trend analysis and multi-dimensional evaluation. Long-term trend analysis is to analyze the dynamically changing security risks by using historical and real-time data in the monitoring area; it analyzes historical data, identifies long-term trends and patterns, and uses patterns in historical data to build prediction models to identify security incidents in advance. The prediction model is expressed as: Where P(y=1X) represents the probability of event y under the condition of feature X, x1, x2, ..., x nrepresents the extracted features, β1,β2,...,β n is the model parameter. By building a prediction model, we can effectively predict the occurrence of intrusion events or other security events and improve the overall security management level. We use historical data to conduct long-term trend analysis and build a prediction model to identify possible security events. The collected data can provide effective security warnings for the monitoring area through intelligent analysis and data reuse technology.

[0060] Multidimensional assessment provides a multidimensional perspective on security risks by integrating data from different monitoring points, time periods, and environmental conditions. After multidimensional analysis, a risk score is assigned to each area and time period, expressed as: R i =w1·F event +w2·F env +w3·F time , where R i represents the risk score of the ith region, F event 、F env 、F time They represent event characteristics, environmental characteristics, and time characteristics, respectively. w1, w2, and w3 represent weight coefficients. High-risk areas and time periods are identified. According to real-time data and historical analysis results, security policies are dynamically adjusted to increase the monitoring frequency of high-risk areas. The application of data reuse technology in dynamic analysis not only provides predictive capabilities and comprehensive evaluation for security management, but also realizes the optimal allocation of resources. The realization of these purposes and functions can greatly improve the efficiency and effectiveness of security management.

[0061] S6. When an abnormal situation is detected, the alarm mechanism is triggered immediately. The monitoring personnel can quickly start the emergency response plan based on the alarm information and conduct remote command and dispatch. When an abnormal situation is detected, the alarm mechanism can be quickly triggered so that the monitoring personnel can receive the alarm in time and handle it accordingly, thus improving the emergency response capability. The monitoring personnel can quickly start the emergency response plan based on the alarm information, conduct remote command and dispatch, and effectively improve the efficiency and success rate of handling emergency situations.

[0062] From the above, we can see that by utilizing the planned monitoring range and number of control balls and making reasonable layouts according to actual needs, we can ensure comprehensive coverage of key areas and improve monitoring efficiency. After integrating the Beidou positioning module and inclination sensor, the control ball can accurately locate and adjust the monitoring angle in real time to ensure the accuracy and timeliness of the monitoring data. At the same time, the transmitted video data can be intelligently analyzed and preprocessed to automatically identify abnormal behaviors, reduce the workload of monitoring personnel, and improve the speed of event response. Through data multiplexing technology, the dynamics in the monitoring area can be analyzed, which can better identify potential threats, optimize resource allocation, and ensure dynamic monitoring of important moments and sites.

[0063] It is important to note that other substitutions, modifications, changes and omissions may be made in the design, operating conditions and arrangement of the exemplary embodiments without departing from the scope of the invention. Therefore, the present invention is not limited to a particular embodiment, but extends to a variety of modifications that still fall within the scope of the appended claims.

[0064] Additionally, in order to provide a concise description of exemplary embodiments, all features of an actual embodiment (ie, those features that are not relevant to the best mode presently contemplated for carrying out the invention or those that are not relevant to implementing the invention) may not be described.

[0065] It should be understood that in the development of any actual implementation, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort may be complex and time-consuming, but for those of ordinary skill having the benefit of this disclosure, without undue experimentation, the development effort will be a routine task of design, fabrication, and production.

[0066] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A security monitoring method based on control ball data multiplexing, characterized in that: The following steps are involved: S1. According to the monitoring requirements and spatial layout, plan the monitoring range and quantity of the control ball, integrate the Beidou positioning module into the control ball, and install the control ball at the predetermined location; S2, using the pan / tilt and tilt sensor to automatically adjust the control ball to a horizontal state; S3, transmit the video and data collected by the control ball to the monitoring center in real time through wireless communication; S4, pre-processing the transmitted video data, performing intelligent analysis on the video data, and storing the processed video and data in a database; S5. Share videos and data with relevant departments, allow multi-terminal access, and use data multiplexing technology to analyze dynamics within the monitoring area; S6. When an abnormal situation is detected, the alarm mechanism is triggered immediately. The monitoring personnel will quickly initiate the emergency response plan based on the alarm information and conduct remote command and dispatch.

2. A security monitoring method based on control ball data multiplexing as claimed in claim 1, characterized in that: The Beidou positioning module is used to record the position information of the control ball in real time, and use the pan / tilt to rotate the control ball horizontally and vertically, and combine with the inclination sensor to detect the inclination angle of the control ball, and automatically adjust to the optimal monitoring state.

3. A security monitoring method based on control ball data multiplexing as claimed in claim 1, characterized in that: The transmitted video data is preprocessed, including denoising, image enhancement and frame extraction; deep features are extracted from the preprocessed image. Assuming that the input image is I, the convolution kernel is K, its size is F×F, and there are C′ filters, the convolution operation is expressed as: Among them, O(i,j) is the value of the output feature map at position (i,j), and (i+m,j+n) represents the movement of the convolution window on the input image; After the convolution operation, the bias is added and processed by a nonlinear activation function, which is expressed as: Z(i,j)=O(i,j)+b A(i,j)=ReLU(Z(i,j)) Among them, Z(i,j) represents the value after adding the bias, b represents the bias term, and A(i,j) represents the output of the activation function; Then perform a pooling operation to reduce the dimension of the feature map, expressed as: Among them, P(i,j) represents the value of the feature map at position (i,j) after pooling, S represents the size of the pooling window, and s is the stride.

4. A security monitoring method based on control ball data multiplexing as claimed in claim 3, characterized in that: Perform anomaly detection analysis on the extracted features to identify abnormal behaviors and learn normal feature representations by reconstructing the input data. X is used as the extracted feature data set, then X = {x1, x2, ..., x n } as feature input, use the encoder to input feature x i Convert to implicit representation h i , and then the implicit representation h i Decode to reconstruct feature x i ′, calculate the reconstruction error for each sample, expressed as: Among them, d represents the dimension of the feature, and MSE represents the mean square error. According to the reconstruction error distribution of the training set, a threshold T is set. When the reconstruction error exceeds this threshold, it is marked as an abnormality. At the same time, a 3D convolutional network is used to directly process the video sequence, capture spatial and temporal features, and identify complex behavior patterns.

5. A security monitoring method based on control ball data multiplexing as claimed in claim 4, characterized in that: Classify detected abnormal behaviors and use trained models to identify and annotate user behaviors, including fighting, stealing, and large-scale gatherings; and use semantic segmentation to distinguish and understand the relationship between pedestrians, vehicles, and background in the scene; The recognition results are combined with predefined security scenario rules to evaluate the recognized events. The rule engine makes judgments based on the rules. After identifying abnormal behaviors, the alarm mechanism is triggered, which automatically generates alarms and notifies monitoring personnel.

6. A security monitoring method based on control ball data multiplexing as claimed in claim 1, characterized in that: Data reuse technology is used to analyze the dynamics within the monitoring area, conduct long-term trend analysis and multi-dimensional assessment. Long-term trend analysis is to analyze dynamically changing security risks by using historical and real-time data within the monitoring area; It analyzes historical data, identifies long-term trends and patterns, and uses the patterns in historical data to build a prediction model to identify security incidents that will occur in advance. The prediction model is expressed as: Where P(y=1X) represents the probability of event y under the condition of feature X, x1, x2, ..., x n represents the extracted features, β1,β2,...,β n is the model parameter.

7. A security monitoring method based on control ball data multiplexing as claimed in claim 6, characterized in that: Multidimensional assessment provides a multidimensional perspective on security risks by integrating data from different monitoring points, time periods, and environmental conditions. After multidimensional analysis, a risk score is assigned to each area and time period, expressed as: R i =w1·F event +w2·F env +w3·F time , where R i represents the risk score of the ith region, F event 、F env 、F time They represent event characteristics, environmental characteristics and time characteristics respectively. w1, w2 and w3 represent weight coefficients. High-risk areas and time periods are identified. According to real-time data and historical analysis results, security policies are dynamically adjusted to increase the monitoring frequency of high-risk areas.

Images