CN118506287A - Regional security monitoring method, system, readable storage medium and computer - Google Patents

Abstract

The present invention provides a regional security monitoring method, system, readable storage medium and computer, the method comprising: controlling a data acquisition device to collect data from a monitoring area under device parameters to obtain grid monitoring data; inputting labeled and pre-processed face image data into a face recognition model for training to obtain a face recognition optimization model; using the face recognition optimization model to perform real-time analysis on the grid monitoring data to extract face feature data, and performing target detection on the monitoring area according to the parameters of each base station to calculate the target location information; performing data fusion on the grid monitoring data, face feature data and target location information, and performing security monitoring on the monitoring area according to the data fusion result. The present invention uses fusion technology to realize automated and intelligent data processing and decision-making, greatly improves monitoring efficiency, reduces manual intervention and response time, reduces labor costs and safety risks, and thus maximizes cost-effectiveness.

Description

Regional security monitoring method, system, readable storage medium and computer

Technical Field

The present invention relates to the field of regional monitoring technology, and in particular to a regional security monitoring method, system, readable storage medium and computer.

Background Art

As security needs continue to increase, single video surveillance, face recognition or positioning technology can no longer meet the needs of high-precision and high-efficiency security monitoring in complex scenarios. Related technologies have technical and methodological deficiencies:

Complexity of data preprocessing: The diversity of data sources leads to a complex preprocessing process, which requires a large amount of information to be processed, poor real-time performance, and high cost. Low-level fusion requires that the original information of the sensor has a high error correction processing capability when fused, which increases the difficulty of technical implementation.

Insufficient complex entity association methods: In cross-language and cross-domain data fusion, unstructured data generally does not explicitly contain attribute names, and its entity attributes do not necessarily appear in structured data, resulting in insufficiencies in the scope of application and accuracy of association methods. Conflict resolution technology relies on the availability of actual reference data, and the lack of domain-specific and targeted reference data narrows its practicality.

Limitations of fusion strategies: Strategies such as statistical-based fusion, model-based fusion, and rule-based fusion all have their limitations and may not be suitable for all types of data and application scenarios. Weight allocation and credibility assessment are often based on experience or heuristic methods, lacking unified evaluation standards and scientific basis.

Limitations of machine learning algorithms: Machine learning algorithms may be affected by outliers and noise during training, resulting in degraded model performance. For high-dimensional, nonlinear data, machine learning algorithms may have difficulty finding effective mapping relationships, resulting in inaccurate fusion results.

How to effectively combine multiple technologies and conduct in-depth data processing to achieve complementary advantages is the current technical challenge.

Summary of the invention

Based on this, the purpose of the present invention is to provide a regional security monitoring method, system, readable storage medium and computer to at least solve the deficiencies in the above-mentioned technology.

The present invention provides a regional security monitoring method, comprising:

Determine the equipment parameters of the data acquisition device based on the area information of the monitoring area, and control the data acquisition device to collect data for the monitoring area under the equipment parameters to obtain corresponding grid monitoring data;

Collecting a number of facial image data, annotating and preprocessing the facial image data, building a facial recognition model using a deep learning algorithm, and inputting the annotated and preprocessed facial image data into the facial recognition model for training to obtain a facial recognition optimization model;

The face recognition optimization model is used to perform real-time analysis on the grid monitoring data to extract corresponding face feature data, base station parameters of each base station are determined according to the regional information, and target detection is performed on the monitoring area according to each base station parameter to calculate corresponding target location information;

The grid monitoring data, the facial feature data and the target location information are fused, and security monitoring of the monitoring area is performed based on the data fusion result.

Further, the step of determining device parameters of a data acquisition device based on the area information of the monitoring area, and controlling the data acquisition device to acquire images of the monitoring area under the device parameters to obtain corresponding gridded monitoring data includes:

Determine the number, type and installation location of the data acquisition equipment based on the area information of the monitoring area, so that the data acquisition equipment covers the monitoring area;

The data acquisition device is used to acquire data from the monitoring area, and the acquired data is converted into a format to obtain corresponding grid monitoring data.

Furthermore, the step of inputting the labeled and preprocessed face image data into the face recognition model for training to obtain the face recognition optimization model includes:

Constructing a face occlusion model and a profile face model, and inputting the labeled and preprocessed face image data into the face occlusion model and the profile face model in sequence to determine the face quality and filter out the profile face image data to obtain preliminary image data;

The preliminary image data is vectorized for facial material, and the vectorization result is compared with the face database, the corresponding face recognition result is identified based on the face confidence threshold, and the face recognition model is trained according to the face recognition result to obtain a face recognition optimization model.

Further, the steps of determining base station parameters of each base station according to the area information, and performing target detection on the monitoring area according to each base station parameter to calculate the corresponding target location information include:

Assigning labels to the targets in the monitoring area, and calculating the flight time and arrival angle of the measurement signal transmitted by each base station to the target between the corresponding base station and the label of the target in the monitoring area;

The distance between each base station and the tag is calculated based on the propagation speed of the measurement signal, the flight time, and the arrival angle, and the target position information of the target is determined based on the distance between each base station and the tag.

Furthermore, the steps of fusing the grid monitoring data, the facial feature data and the target location information, and performing security monitoring on the monitoring area according to the data fusion result include:

Preprocessing the gridded monitoring data, the facial feature data, and the target location information, and extracting features based on the data preprocessing results to obtain corresponding feature data;

The feature data is subjected to feature fusion, and the fused features are subjected to decision classification to obtain corresponding fused data, and security monitoring of the monitoring area is performed based on the fused data.

The present invention also provides a regional security monitoring system, comprising:

A data acquisition module, used to determine the equipment parameters of the data acquisition device based on the area information of the monitoring area, and control the data acquisition device to collect data for the monitoring area under the equipment parameters to obtain corresponding grid monitoring data;

A model optimization module is used to collect a number of facial image data, annotate and preprocess the facial image data, build a facial recognition model using a deep learning algorithm, and input the annotated and preprocessed facial image data into the facial recognition model for training to obtain a facial recognition optimization model;

A position information calculation module, used to use the face recognition optimization model to perform real-time analysis on the grid monitoring data to extract corresponding face feature data, determine base station parameters of each base station according to the area information, and perform target detection on the monitoring area according to each base station parameter to calculate corresponding target position information;

The security monitoring module is used to fuse the grid monitoring data, the facial feature data and the target location information, and perform security monitoring on the monitoring area according to the data fusion result.

Furthermore, the data acquisition module includes:

A parameter determination unit, configured to determine the number, type and installation location of the data acquisition equipment based on the area information of the monitoring area, so that the data acquisition equipment covers the monitoring area;

The data collection unit is used to collect data from the monitoring area using the data collection device, and convert the format of the collected data to obtain corresponding grid monitoring data.

Furthermore, the model optimization module includes:

An image processing unit, used to construct a face occlusion model and a profile face model, and input the labeled and pre-processed face image data into the face occlusion model and the profile face model in sequence to determine the face quality and filter out the profile face image data to obtain preliminary image data;

The model optimization unit is used to vectorize the face material of the preliminary image data, compare the vectorization result with the face database, identify the corresponding face recognition result based on the face confidence threshold, and train the face recognition model according to the face recognition result to obtain a face recognition optimization model.

Furthermore, the position information calculation module includes:

A label assignment unit, configured to assign labels to the targets in the monitoring area, and calculate the flight time and arrival angle of the measurement signal transmitted by each base station to the target between the corresponding base station and the label of the target in the monitoring area;

The position information calculation unit is used to calculate the distance between each base station and the tag based on the propagation speed of the measurement signal, the flight time and the arrival angle, and determine the target position information of the target based on the distance between each base station and the tag.

Furthermore, the security monitoring module includes:

A data preprocessing unit, used for preprocessing the grid monitoring data, the face feature data and the target location information, and extracting features according to the data preprocessing results to obtain corresponding feature data;

The security monitoring unit is used to perform feature fusion on the feature data, and make decision classification on the fused features to obtain corresponding fused data, and perform security monitoring on the monitoring area according to the fused data.

The present invention also provides a readable storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned regional safety monitoring method is implemented.

The present invention also proposes a computer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned regional safety monitoring method when executing the computer program.

The regional security monitoring method, system, readable storage medium and computer of the present invention determine the equipment parameters of the data acquisition device based on the regional information of the monitoring area, use the data acquisition device to collect data on the monitoring area to obtain grid monitoring data, provide comprehensive visual information for subsequent data fusion, use the face recognition model to perform real-time analysis on the grid monitoring data to extract facial feature data, so as to accurately identify the target identity, determine the base station parameters of each base station based on the regional information, and perform target detection on the monitoring area based on the parameters of each base station to calculate the target position information, accurately track the target position, fuse the grid monitoring data, facial feature data and target position information, and perform security monitoring on the monitoring area based on the data fusion result, use fusion technology to realize automatic and intelligent data processing and decision-making, greatly improve monitoring efficiency, reduce manual intervention and response time, reduce labor costs and safety risks, and thus maximize cost-effectiveness.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a flow chart of a regional security monitoring method according to a first embodiment of the present invention;

FIG2 is a detailed flow chart of step S101 in FIG1 ;

FIG3 is a detailed flow chart of step S102 in FIG1 ;

FIG4 is a detailed flow chart of step S103 in FIG1 ;

FIG5 is a detailed flow chart of step S104 in FIG1 ;

FIG6 is a block diagram of a regional security monitoring system according to a second embodiment of the present invention;

FIG. 7 is a block diagram of the structure of a computer in the third embodiment of the present invention.

The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings.

DETAILED DESCRIPTION

In order to facilitate understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and/or" used herein includes any and all combinations of one or more of the related listed items.

Embodiment 1

Please refer to FIG. 1 , which shows a method for regional security monitoring in a first embodiment of the present invention. The method specifically includes steps S101 to S104:

S101, determining device parameters of a data acquisition device based on area information of a monitoring area, and controlling the data acquisition device to collect data for the monitoring area under the device parameters to obtain corresponding grid monitoring data;

Further, referring to FIG. 2 , the step S101 specifically includes steps S1011 to S1012:

S1011, determining the number, type and installation location of data acquisition devices based on the area information of the monitoring area, so that the data acquisition devices cover the monitoring area;

S1012: Utilize the data acquisition device to collect data from the monitoring area, and convert the format of the collected data to obtain corresponding grid monitoring data.

In the specific implementation, the number, type and installation location of the required data acquisition devices (in this embodiment, the data acquisition devices are cameras) are determined according to the size, shape and key areas of the monitoring area. Ensure that the camera's field of view can cover the entire monitoring area while avoiding too many overlapping areas. Associate and match the camera with the segmented grid pieces.

Specifically, the camera needs to effectively identify human faces, and the optimal identification distance is within 10 meters. The on-site environment is divided and laid out in a geometric grid of 10X10 meters, and the grid numbers after division are marked, such as: 1-1, 1-2, 1-3. The purpose of the monitoring area and the requirements for the monitoring screen are clearly defined. The camera resolution should be above 1920X1080P, and the field of view should be controlled within the most effective identification range of 10 meters. Infrared monitoring equipment should be added in dimly lit locations, and the video equipment should be installed at a height of 3.5 meters.

Choose the right camera type based on the monitoring area and monitoring needs. For example, for short-distance monitoring, you can choose a dome camera or an infrared integrated camera; for medium and long-distance monitoring, a gun camera or a ball camera may be more suitable. Calculate the number of cameras required based on the area and shape of the monitoring area and the camera's field of view; determine the best installation location for each camera based on the shape of the monitoring area and the camera's field of view. This usually involves precise measurement and planning of the installation location to ensure that the camera's field of view can cover the entire monitoring area while avoiding too much overlapping area.

Furthermore, according to the monitoring area and the optimal monitoring range of the camera, bind the monitoring grid number set that each camera is responsible for. Install the camera according to the predetermined installation location. This includes fixing the camera bracket, connecting the power cable and data cable, etc. During the installation process, it is necessary to ensure the stability and safety of the camera to avoid accidents. After the installation is completed, configure and test the monitoring system. This includes setting the camera parameters (such as resolution, frame rate, etc.), adjusting the camera's viewing angle and focal length, and testing the system's stability and picture quality.

Specifically, after the camera is deployed, a video stream transmission network is deployed to ensure that the camera can stably and efficiently transmit the video stream to the central server. Configure the video streaming platform service on the central server to register the device according to the national standard 28181, and perform operations such as decoding, format conversion and storage on the received video stream. Output into various video protocol streams (rtsp/rtmp/HLS/flv/webrtc, etc.) for the front-end terminal to play. Video acquisition process: ① Monitoring terminal: The monitoring terminal is mainly a video device or a device that provides supporting video storage. Its main function is to provide video streams (national standard cameras need to be registered to the streaming media server). ② Streaming media platform: The streaming media platform mainly includes SIP signaling authentication, and provides a series of camera services such as national standard camera registration, deregistration, heartbeat, keep alive, open stream, close stream, etc.; various video streaming protocol push-pull stream live broadcast on demand and other services; ③ Display terminal: Mainly for the client application end (mobile phone, computer, electronic wall, etc.), receiving the video stream encoded and decoded by streaming media for display and playback.

S102, collecting a number of facial image data, annotating and preprocessing the facial image data, building a facial recognition model using a deep learning algorithm, and inputting the annotated and preprocessed facial image data into the facial recognition model for training to obtain a facial recognition optimization model;

Further, referring to FIG. 3 , the step S102 specifically includes steps S1021 to S1022:

S1021, constructing a face occlusion model and a profile face model, and inputting the labeled and preprocessed face image data into the face occlusion model and the profile face model in sequence to determine the face quality and filter out the profile face image data to obtain preliminary image data;

S1022, vectorize the face material of the preliminary image data, and compare the vectorization result with the face database, identify the corresponding face recognition result based on the face confidence threshold, and train the face recognition model according to the face recognition result to obtain a face recognition optimization model.

In the specific implementation, a large amount of facial image data is collected, annotated and preprocessed, and a deep learning algorithm (such as convolutional neural network) is used to build a face recognition model, and the annotated data is used for training. In order to improve the recognition accuracy, the face occlusion model and the profile face model are trained. Non-frontal or occluded faces are filtered to improve accuracy.

The trained face recognition model is deployed on the central server, and the above-obtained video stream (grid monitoring data) is analyzed in real time to extract facial features. The extracted facial features are compared with the pre-stored face database to achieve identity authentication.

S103, using the face recognition optimization model to perform real-time analysis on the grid monitoring data to extract corresponding face feature data, determine base station parameters of each base station according to the area information, and perform target detection on the monitoring area according to each base station parameter to calculate corresponding target location information;

Further, referring to FIG. 4 , the step S103 specifically includes steps S1031 to S1032:

S1031, assigning tags to the targets in the monitoring area, and calculating the flight time and arrival angle of the measurement signal transmitted by each base station to the target between the corresponding base station and the tag of the target in the monitoring area;

S1032: Calculate the distance between each base station and the tag based on the propagation speed of the measurement signal, the flight time, and the arrival angle, and determine the target position information of the target based on the distance between each base station and the tag.

In specific implementation, the face recognition process is as follows:

① Video stream access: The monitoring terminal or video streaming media provides video streams using the RTSP protocol.

② Video processing unit: The data processing unit accesses the standard video stream and extracts video frame images according to the configured recognition speed parameters and rules.

③ Face occlusion judgment: Based on the occlusion model obtained after the previous occlusion training, this model mainly judges the quality of the face and obtains the proportion of the face occluded by obstacles. If the proportion of the face occluded by obstacles exceeds the threshold, the next step of face recognition will not be carried out.

④ Side face judgment: mainly based on the side face model made in the early stage, judge the face quality and whether it is a side face picture (the side face judgment rule is: whether the eye distance meets the threshold setting requirements). If it is a side face picture, the next step of face recognition business will not be continued. Only the front face picture will proceed to the next step of face recognition logic.

Eye distance judgment: Get the key points of the face and calculate whether the distance between the two eyes is greater than the distance of the set ratio of the face. If it is less, filter it.

The calculation formula is as follows:

;

Wherein, x ₁ , y ₁ represent the coordinates of the left eye relative to the pixel of the picture; x ₂ , y ₂ represent the coordinates of the right eye relative to the pixel of the picture, and |AB| represents the distance between the left and right eyes;

⑤ Face recognition: First, face quality is judged (det_score>=threshold). If the threshold is passed, it means that the relevant picture contains a face material that meets the requirements. Pictures that do not meet the face quality requirements will exit the subsequent judgment logic. The face material of the face photo that meets the requirements will be vectorized and compared with the data in the face database. The comparison result is the face confidence. By judging whether the face confidence value exceeds the face comparison threshold, the relevant results will be output after the comparison is completed.

The structure of successful face recognition is as follows:

{

"content": "Face Recognition",#Algorithm Type Name

"data":

{

"object": "Face recognition",#Algorithm type name

"type_id": 109,#algorithm type code

"det_score": 0.8930664 #Is it a face probability?

"face_res": #Recognize face information

{

"user_id": "6f4466c4753a48049348c0ccf2f85c8d", #Face ID

"confidence": 0.7826316709641026, #face confidence

"name": "朱宇", #face name

}

}

};

If the comparison fails, it means it is a stranger. The structure is as follows:

{

"content": "Face Recognition",#Algorithm Type Name

"data":

{

"object": "Face recognition",#Algorithm type name

"type_id": 109,#algorithm type code

"det_score": 0.8230664 #Is it a face probability?

"face_res": #Recognize face information

{

"user_id": "unknown", #Face ID

"confidence": 0.4826316709641026, #face confidence

"name": "Stranger", #face name

}

}

};

⑥Result output: Give the userId and type_id in the face recognition result to the business service for the next step.

Specifically, according to the size and shape of the monitoring area, the relevant base stations are deployed as follows:

Corridor-like long channels: Deploy a positioning base station every 10 meters according to the channel length to form a two-dimensional base station group. Room-like rectangular areas are covered according to a 5-meter radius and base station coverage is planned to form a three-dimensional base station group.

Furthermore, the target that needs to be located (such as bid evaluation experts) is equipped with a UWB tag, and the location information of the target is calculated by measuring the flight time (ToF) and arrival angle (AoA) of the UWB signal between the base station and the tag.

TOF (Time of Flight) Flight Time Method:

1. Signal sending and receiving:

The UWB base station (or anchor point) sends UWB signals.

When the target tag (or mobile device) receives the signal, it records the time of reception and immediately sends a response signal back to the base station.

After receiving the response signal, the base station records the receiving time.

2. Time measurement:

Calculate the round-trip time (RTT) of the signal from the base station to the tag. RTT is the time it takes for the response signal to return to the base station minus the time it takes for the signal to be sent from the base station.

Since RTT is the total time it takes for the signal to travel back and forth, it needs to be divided by 2 to get the one-way time of flight (TTOF).

3. Distance calculation:

Knowing the propagation speed of the signal in the air (close to the speed of light c), the distance from the base station to the tag can be calculated using the formula distance = TTOF * c.

4. Location determination:

If there are three or more base stations in communication with the tag, and the locations of those base stations are known, then trilateration (or multilateration) can be used to determine the tag's location.

Draw a circle with each base station as the center and the distance from the base station to the tag as the radius. The intersection of these circles is the location of the tag.

AOA (Angle of Arrival) arrival angle method:

1. Signal reception and angle measurement:

Multiple antennas or array antennas are configured on the base station to receive the signals sent by the tags.

By measuring the time difference or phase difference when the signal arrives at different antennas, the angle at which the signal arrives at the base station can be calculated.

2. Angle determination:

Specific algorithms (such as phase interferometry, beamforming, etc.) are used to determine the precise angle at which the signal arrives at the base station.

3. Location determination:

If the precise locations of two or more base stations and the angles of arrival of the signals they each receive are known, the location of the tag can be determined by solving the geometric relationship of the intersection of the angles.

It should be noted that the AOA method usually requires at least two non-collinear base stations to accurately determine the location of the tag.

Specifically, UWB positioning correction optimization algorithms usually include a series of technologies and methods to reduce positioning errors and improve positioning accuracy. These algorithms may involve processing raw measurement data, improving communication between base stations and target tags, and optimizing positioning calculation methods.

1. Data preprocessing algorithm

Filtering algorithms: such as Kalman filtering and extended Kalman filtering (EKF), are used to smooth the measurement data and reduce the impact of noise and random errors.

Error correction algorithms: Methods based on statistical models or machine learning to identify and correct systematic errors.

2. Communication Optimization Algorithm

Signal enhancement algorithms: such as beamforming and multiple-input multiple-output (MIMO) technology, are used to improve signal transmission quality and coverage.

Time synchronization algorithm: Ensure time synchronization between base stations and between base stations and tags to reduce positioning errors caused by clock differences.

3. Positioning calculation optimization algorithm

Multi-base station positioning algorithms: such as the least squares method and the weighted least squares method, improve positioning accuracy by combining the measurement data of multiple base stations.

Model-based methods: such as using geometric relationships, signal propagation models, etc. to perform positioning calculations.

Optimization process:

The optimization process of UWB positioning system usually includes the following steps:

1. Data collection: Collect a large amount of positioning data in actual scenarios, including the coordinates of base stations and tags, measured distances, timestamps, etc.

2. Data analysis: Analyze the collected data to identify the main sources and characteristics of positioning errors.

3. Algorithm selection and design: According to the data analysis results, select the appropriate correction optimization algorithm and make necessary algorithm design and adjustments.

4. Experimental verification: Conduct experimental verification in actual scenarios to evaluate the effect and performance of the optimization algorithm.

5. Iterative optimization: Based on the experimental results, the algorithm is iteratively optimized until the expected performance indicators are met.

Corresponding compensation:

In UWB positioning systems, compensation is usually used to reduce or eliminate positioning errors caused by various factors. The compensation methods vary depending on the system design and application scenario, but generally include the following:

1. Hardware compensation: Reduce hardware errors by improving hardware design and using higher-precision measuring equipment.

2. Software compensation: The measured data is processed, corrected and compensated through algorithms to reduce errors in software algorithms and data processing.

3. Environmental compensation: Compensate for errors under specific environmental conditions, such as multipath effects, non-line-of-sight (NLOS) errors, etc. This may require the establishment of environmental models and the use of machine learning and other methods for prediction and compensation.

4. Clock compensation: For time synchronization errors, a clock compensation algorithm can be used to correct the time differences between base stations and between base stations and tags.

S104, fusing the grid monitoring data, the facial feature data and the target location information, and performing security monitoring on the monitoring area according to the data fusion result.

Further, referring to FIG. 5 , the step S104 specifically includes steps S1041 to S1042:

S1041, performing data preprocessing on the grid monitoring data, the facial feature data, and the target location information, and performing feature extraction according to the data preprocessing result to obtain corresponding feature data;

S1042, performing feature fusion on the feature data, and performing decision classification on the fused features to obtain corresponding fused data, and performing security monitoring on the monitoring area according to the fused data.

In the specific implementation, the above-mentioned grid monitoring data, facial feature data and target location information are received, and the received data is time synchronized to ensure the consistency of data from different sources in time.

Design a suitable data fusion algorithm to fuse video stream data, identity authentication results and positioning data, set the weight and credibility of data sources, and the correlation between them to improve the accuracy of fusion results.

1. The steps of data fusion are as follows:

1. Data preprocessing:

Data cleaning: Remove duplicate, erroneous or invalid data.

Data conversion: Convert data from different sources into a unified format and structure for subsequent processing.

2. Data Fusion:

Feature extraction: Extract useful features from preprocessed data. These features may include shape, color, texture, position, etc.

Feature fusion: Fusion of features from different data sources to form a unified feature representation. This can be achieved through simple weighted averaging, principal component analysis, neural networks, etc.

Decision fusion: Based on feature fusion, decisions or classifications are made based on the fused features. This usually involves one or more classifiers or regression models.

3. Weight and credibility allocation:

Weight assignment: The data quality and importance of different data sources may be different, so different weights need to be assigned to them. The weights can be determined by experience, expert knowledge, or data quality assessment. During the fusion process, the weights can be used to adjust the contribution of different data sources to the final result.

Credibility assessment: Credibility reflects the accuracy and reliability of data. It can be evaluated through a variety of methods, such as statistical methods based on historical data, model-based methods, etc. In data fusion, credibility can be used to filter out low-quality data or adjust the weight distribution in the fusion process.

4. Result output and evaluation:

The fused data or decision results need to be output in an appropriate manner for the convenience of users or subsequent processing.

Evaluate fusion effect: Evaluate the performance and effect of the fusion algorithm by comparing the fusion result with the actual result or reference result. Evaluation indicators may include accuracy, completeness, consistency, etc.

2. Specific treatment of weight and credibility

1. Weight processing:

In the feature fusion stage, a weight value can be assigned to the features of each data source, which reflects the importance or contribution of the data source.

In the decision fusion stage, the output of each classifier or regression model can be assigned a weight value that reflects the performance or reliability of the model in a specific task.

Weight values can be adjusted automatically through learning algorithms (such as weight update algorithms in machine learning) or can be set manually through expert knowledge or experience.

2. Credibility processing:

During the data preprocessing stage, the credibility of the data can be preliminarily determined through data quality assessment.

During the fusion process, the credibility of the data can be dynamically adjusted according to the actual performance of the data (such as classification accuracy, prediction error, etc.).

For data with low credibility, you can reduce its weight, filter it out, or perform special processing on it.

Specifically, the fusion processing results are pushed to the corresponding security monitoring and early warning module through the API interface service. According to the fusion results, early warnings are issued for the generated alarm information. The early warning methods include sending system station messages and the twin platform flashing highlights.

The security monitoring and early warning module is implemented in the following ways:

1. Abnormal behavior detection and identification

Use machine learning algorithms to analyze and learn the fused data, identify abnormal behavior patterns, monitor the target's behavior in real time, and trigger an alarm mechanism when abnormal behavior is detected.

2. Design and implementation of alarm mechanism

Design appropriate alarm mechanisms, including sound alarms, SMS notifications, email notifications, etc., and select appropriate alarm methods and alarm levels based on the severity and urgency of abnormal behavior.

3. Alarm information processing and feedback

The triggered alarm information is recorded and processed, including saving alarm videos, generating alarm reports, etc., and the alarm information is promptly fed back to relevant personnel for processing and response to ensure timely handling of security incidents.

In summary, the regional security monitoring method in the above-mentioned embodiment of the present invention determines the equipment parameters of the data acquisition device based on the regional information of the monitoring area, uses the data acquisition device to collect data on the monitoring area to obtain grid monitoring data, provides comprehensive visual information for subsequent data fusion, and uses a face recognition model to perform real-time analysis on the grid monitoring data to extract facial feature data to accurately identify the target identity, determines the base station parameters of each base station based on the regional information, and performs target detection on the monitoring area based on the parameters of each base station to calculate the target position information, accurately tracks the target position, fuses the grid monitoring data, facial feature data and target position information, and performs security monitoring on the monitoring area based on the data fusion result, and uses fusion technology to realize automated and intelligent data processing and decision-making, greatly improves monitoring efficiency, reduces manual intervention and response time, reduces labor costs and safety risks, and thus maximizes cost-effectiveness.

Embodiment 2

Another aspect of the present invention further provides a regional security monitoring system. Please refer to FIG6 , which shows a regional security monitoring system in a second embodiment of the present invention. The system includes:

The data acquisition module 11 is used to determine the equipment parameters of the data acquisition device based on the area information of the monitoring area, and control the data acquisition device to collect data for the monitoring area under the equipment parameters to obtain corresponding grid monitoring data;

Furthermore, the data acquisition module 11 includes:

A parameter determination unit, configured to determine the number, type and installation location of the data acquisition equipment based on the area information of the monitoring area, so that the data acquisition equipment covers the monitoring area;

The data collection unit is used to collect data from the monitoring area using the data collection device, and convert the format of the collected data to obtain corresponding grid monitoring data.

The model optimization module 12 is used to collect a number of facial image data, annotate and preprocess the facial image data, build a facial recognition model using a deep learning algorithm, and input the annotated and preprocessed facial image data into the facial recognition model for training to obtain a facial recognition optimization model;

Furthermore, the model optimization module 12 includes:

An image processing unit, used to construct a face occlusion model and a profile face model, and input the labeled and pre-processed face image data into the face occlusion model and the profile face model in sequence to determine the face quality and filter out the profile face image data to obtain preliminary image data;

The model optimization unit is used to vectorize the face material of the preliminary image data, compare the vectorization result with the face database, identify the corresponding face recognition result based on the face confidence threshold, and train the face recognition model according to the face recognition result to obtain a face recognition optimization model.

The position information calculation module 13 is used to perform real-time analysis on the grid monitoring data using the face recognition optimization model to extract corresponding face feature data, determine the base station parameters of each base station according to the area information, and perform target detection on the monitoring area according to each of the base station parameters to calculate the corresponding target position information;

Furthermore, the position information calculation module 13 includes:

A label assignment unit, configured to assign labels to the targets in the monitoring area, and calculate the flight time and arrival angle of the measurement signal transmitted by each base station to the target between the corresponding base station and the label of the target in the monitoring area;

The position information calculation unit is used to calculate the distance between each base station and the tag based on the propagation speed of the measurement signal, the flight time and the arrival angle, and determine the target position information of the target based on the distance between each base station and the tag.

The security monitoring module 14 is used to fuse the grid monitoring data, the facial feature data and the target location information, and perform security monitoring on the monitoring area according to the data fusion result.

Furthermore, the security monitoring module 14 includes:

A data preprocessing unit, used for preprocessing the grid monitoring data, the face feature data and the target location information, and extracting features according to the data preprocessing results to obtain corresponding feature data;

The security monitoring unit is used to perform feature fusion on the feature data, and make decision classification on the fused features to obtain corresponding fused data, and perform security monitoring on the monitoring area according to the fused data.

The functions or operation steps implemented when the above modules and units are executed are generally the same as those in the above method embodiments, and will not be repeated here.

The regional security monitoring system provided in the embodiment of the present invention has the same implementation principle and technical effects as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the system embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.

Embodiment 3

The present invention also proposes a computer, please refer to Figure 7, which is a computer in the third embodiment of the present invention, including a memory 10, a processor 20, and a computer program 30 stored in the memory 10 and executable on the processor 20. When the processor 20 executes the computer program 30, the above-mentioned regional security monitoring method is implemented.

The memory 10 includes at least one type of readable storage medium, which includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 10 may be an internal storage unit of a computer, such as a hard disk of the computer. In other embodiments, the memory 10 may also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card, etc. Further, the memory 10 may also include both an internal storage unit of the computer and an external storage device. The memory 10 may be used not only to store application software and various types of data installed in the computer, but also to temporarily store data that has been output or is to be output.

Among them, in some embodiments, the processor 20 can be an electronic control unit (Electronic Control Unit, abbreviated as ECU, also known as a vehicle computer), a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, a microprocessor or other data processing chip, used to run the program code stored in the memory 10 or process data, such as executing access restriction programs.

It should be noted that the structure shown in FIG. 7 does not constitute a limitation on the computer. In other embodiments, the computer may include fewer or more components than shown in the figure, or a combination of certain components, or a different arrangement of components.

The embodiment of the present invention further provides a readable storage medium on which a computer program is stored. When the program is executed by a processor, the regional security monitoring method as described above is implemented.

Those skilled in the art will appreciate that the logic and/or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, "computer-readable medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims (10)

1. A regional security monitoring method, characterized by comprising: Determine the equipment parameters of the data acquisition device based on the area information of the monitoring area, and control the data acquisition device to collect data for the monitoring area under the equipment parameters to obtain corresponding grid monitoring data; Collecting a number of facial image data, annotating and preprocessing the facial image data, building a facial recognition model using a deep learning algorithm, and inputting the annotated and preprocessed facial image data into the facial recognition model for training to obtain a facial recognition optimization model; The face recognition optimization model is used to perform real-time analysis on the grid monitoring data to extract corresponding face feature data, base station parameters of each base station are determined according to the regional information, and target detection is performed on the monitoring area according to each base station parameter to calculate corresponding target location information; The grid monitoring data, the facial feature data and the target location information are fused, and security monitoring of the monitoring area is performed based on the data fusion result. 2. The regional security monitoring method according to claim 1 is characterized in that the step of determining the equipment parameters of the data acquisition device based on the regional information of the monitoring area, and controlling the data acquisition device to perform image acquisition on the monitoring area under the equipment parameters to obtain the corresponding grid monitoring data comprises: Determine the number, type and installation location of the data acquisition equipment based on the area information of the monitoring area, so that the data acquisition equipment covers the monitoring area; The data acquisition device is used to acquire data from the monitoring area, and the acquired data is converted into a format to obtain corresponding grid monitoring data. 3. The regional security monitoring method according to claim 1, characterized in that the step of inputting the labeled and pre-processed face image data into the face recognition model for training to obtain the face recognition optimization model comprises: Constructing a face occlusion model and a profile face model, and inputting the labeled and preprocessed face image data into the face occlusion model and the profile face model in sequence to determine the face quality and filter out the profile face image data to obtain preliminary image data; The preliminary image data is vectorized for facial material, and the vectorization result is compared with the face database, the corresponding face recognition result is identified based on the face confidence threshold, and the face recognition model is trained according to the face recognition result to obtain a face recognition optimization model. 4. The regional security monitoring method according to claim 1, characterized in that the steps of determining base station parameters of each base station according to the regional information, and performing target detection on the monitoring area according to each base station parameter to calculate the corresponding target location information include: Assigning labels to the targets in the monitoring area, and calculating the flight time and arrival angle of the measurement signal transmitted by each base station to the target between the corresponding base station and the label of the target in the monitoring area; The distance between each base station and the tag is calculated based on the propagation speed of the measurement signal, the flight time, and the arrival angle, and the target position information of the target is determined based on the distance between each base station and the tag. 5. The regional security monitoring method according to claim 1, characterized in that the step of fusing the grid monitoring data, the facial feature data and the target location information, and performing security monitoring on the monitoring area according to the data fusion result comprises: Preprocessing the gridded monitoring data, the facial feature data, and the target location information, and extracting features based on the data preprocessing results to obtain corresponding feature data; The feature data is subjected to feature fusion, and the fused features are subjected to decision classification to obtain corresponding fused data, and security monitoring of the monitoring area is performed based on the fused data. 6. A regional security monitoring system, characterized by comprising: A data acquisition module, used to determine the equipment parameters of the data acquisition device based on the area information of the monitoring area, and control the data acquisition device to collect data for the monitoring area under the equipment parameters to obtain corresponding grid monitoring data; A model optimization module is used to collect a number of facial image data, annotate and preprocess the facial image data, build a facial recognition model using a deep learning algorithm, and input the annotated and preprocessed facial image data into the facial recognition model for training to obtain a facial recognition optimization model; A position information calculation module, used to use the face recognition optimization model to perform real-time analysis on the grid monitoring data to extract corresponding face feature data, determine base station parameters of each base station according to the area information, and perform target detection on the monitoring area according to each base station parameter to calculate corresponding target position information; The security monitoring module is used to fuse the grid monitoring data, the facial feature data and the target location information, and perform security monitoring on the monitoring area according to the data fusion result. 7. The regional security monitoring system according to claim 6, characterized in that the data acquisition module comprises: A parameter determination unit, configured to determine the number, type and installation location of the data acquisition equipment based on the area information of the monitoring area, so that the data acquisition equipment covers the monitoring area; The data collection unit is used to collect data from the monitoring area using the data collection device, and convert the format of the collected data to obtain corresponding grid monitoring data. 8. The regional security monitoring system according to claim 6, wherein the model optimization module comprises: An image processing unit, used to construct a face occlusion model and a profile face model, and input the labeled and pre-processed face image data into the face occlusion model and the profile face model in sequence to determine the face quality and filter out the profile face image data to obtain preliminary image data; The model optimization unit is used to vectorize the face material of the preliminary image data, compare the vectorization result with the face database, identify the corresponding face recognition result based on the face confidence threshold, and train the face recognition model according to the face recognition result to obtain a face recognition optimization model. 9. A readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the method for regional safety monitoring as claimed in any one of claims 1 to 5 is implemented. 10. A computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the regional security monitoring method as claimed in any one of claims 1 to 5 when executing the computer program.