Community intelligent security alarm method and system based on big data

Abstract

The invention relates to the field of security and protection, in particular to a community intelligent security and protection alarm method and system based on big data. The method comprises the following steps: acquiring access control identification data and monitoring capture data, setting a community database, and constructing a community security network according to the community database, the access control identification data and the monitoring capture data; analyzing the access control identification data and the monitoring capture data through a community security network to obtain a face vector and a verification group vector, and analyzing the face vector and the verification group vector to obtain a tracking value; obtaining a behavior process line according to the community security network and the face vector, and obtaining a behavior value and a behavior comprehensive score according to the behavior process line and the tracking value; and obtaining a recording result through the behavior comprehensive score and the behavior value, and generating an alarm signal and a security strategy according to the obtained recording result. The community management level can be improved.

Description

Technical Field

[0001] The present invention relates to the field of security, and specifically relates to a community intelligent security alarm method and system based on big data. Background Art

[0002] With the acceleration of the urbanization process and the continuous increase in population density, as the basic unit of urban governance, the public security issues of communities have gradually attracted wide social attention; with the rapid development of big data, artificial intelligence, Internet of Things, and cloud computing technologies, the "intelligent security" system driven by data and supported by models has gradually become a new direction for community governance.

[0003] All kinds of security devices deployed in communities (such as cameras, access control, sensors, etc.) often have isolated information and lack the ability of unified platform integration and intelligent analysis; the efficiency of detecting abnormal behaviors of residents is low, reducing the community security guarantee, and the monitoring and analysis efficiency of people suddenly appearing in the community is low; the manual security method is not only costly but also inefficient; traditional cameras only record events and cannot automatically detect and handle security threats, which are the problems we need to solve. Summary of the Invention

[0004] The purpose of the present invention is to propose a community intelligent security alarm method based on big data for the problems existing in the background art.

[0005] The technical solution of the present invention: A community intelligent security alarm method based on big data includes the following steps:

[0006] S1. Obtain access control recognition data and monitoring capture data, set up a community database, and construct a community security network according to the community database, access control recognition data, and monitoring capture data;

[0007] S2. Analyze the access control recognition data and monitoring capture data through the community security network to obtain face vectors and verification group vectors, and analyze the face vectors and verification group vectors to obtain tracking values;

[0008] S3. Obtain a behavior history line according to the community security network and face vectors, and obtain a behavior value and a comprehensive behavior score according to the behavior history line and tracking values;

[0009] S4. Obtain a recording result through the comprehensive behavior score and behavior value, and generate an alarm signal and a security strategy according to the obtained recording result.

[0010] Preferably, the process of obtaining access control recognition data and monitoring capture data, setting up a community database, and constructing a community security network according to the community database, access control recognition data, and monitoring capture data includes:

[0011] The access control recognition data includes the residential information, fingerprint data, and face image data of residents; the monitoring capture data includes the resident image data, camera location data, and capture time;

[0012] Set up a community database; the community database includes verified face image data, verified fingerprint data, valid residential information, residential related relationships, and a community map;

[0013] Analyze the community database, access control recognition data, and monitoring capture data to obtain monitored residents;

[0014] Set up monitoring data points and access control points. Combining with GIS technology, according to the camera location data, monitoring data points, access control points, and community map, construct data interaction relationships and association relationships. According to the data interaction relationships, association relationships, monitoring data points, and access control points, obtain a monitoring network. Store the capture time and resident image data into the corresponding monitoring data points, and store the residential information, fingerprint data, and face image data corresponding to the monitored residents into the access control points to obtain a community security network.

[0015] Preferably, the process of analyzing the community database, access control recognition data, and monitoring capture data to obtain monitored residents includes:

[0016] According to the community database and access control recognition data, obtain valid residential information consistent with the residential information;

[0017] If there is valid residential information consistent with the residential information in the community database, obtain the residential related relationship of the valid residential information, and through the residential related relationship, obtain the corresponding verified face image data and verified fingerprint data. Through face verification technology and fingerprint verification technology, compare the verified face image data and verified fingerprint data with the face image data and fingerprint data respectively. If the verified face image data is corresponding and consistent with the face image data or the verified fingerprint data is corresponding and consistent with the fingerprint data, the resident corresponding to the residential information is a resident of this community, passes through the access control, and record the resident corresponding to the residential information as a monitored resident.

[0018] Preferably, through the community security network, analyze the access control recognition data and monitoring capture data to obtain a face vector and a verification group vector, and the process of analyzing the face vector and the verification group vector to obtain a tracking value is:

[0019] Convert the resident image data in the monitoring data points of the community security network and the verified face image data in the access control points into grayscale images. Through face detection technology, extract the face area in the grayscale images to obtain a monitored face image and a verified face image. According to the verified face image, obtain a verified face feature image;

[0020] Set up a face recognition model; input the monitored face image and the verified face feature image into the face recognition model to obtain a face vector and a verification vector, and obtain a verification group vector based on the verified face image and the verification vector;

[0021] Set up a face encoding; obtain a verification marker vector and a verification marker group vector based on the face encoding, the verification group vector, and the verification vector; through the community security network, match each item of the verification marker vector of the verification marker group vector with the face vector, and the matching specifically means making the verification marker vector of the verification marker group vector and the face vector have as many consecutive equal items as possible; count the number of consecutive equal items, the number of vectors, and the total number of items of the verification marker group vector between the face vector and the verification marker group vector; obtain a tracking value of the face vector based on the number of consecutive equal items, the number of vectors, and the total number of items of the verification marker group vector between the face vector and the verification marker group vector;

[0022] Set up a tracking threshold;

[0023] When the tracking value of the face vector is greater than or equal to the tracking threshold, obtain the last item of the verification marker vector group and insert the obtained last item into the last item of the face vector.

[0024] Preferably, the process of obtaining the behavior history line based on the community security network and the face vector includes:

[0025] Through the community security network, obtain the last item of the face vector of each monitoring data point, denoted as the tracking item, and link the monitoring data points corresponding to the face vectors with the same tracking item and the corresponding resident image data in the order of capture time to obtain the behavior history line of the resident.

[0026] Preferably, the process of obtaining the behavior value and the comprehensive behavior score based on the behavior history line and the tracking value includes:

[0027] Obtain the number of resident images in each monitoring data point of the behavior history line, and obtain the behavior value based on the number of resident images;

[0028] Set up a behavior standard value; when the behavior value is greater than or equal to the behavior standard value, mark the monitoring data point corresponding to the resident image data as an abnormal point; when the behavior value is less than the behavior standard value, mark the monitoring data point corresponding to the resident image data as a normal point; count the number of normal points and the total number of abnormal points and normal points, and obtain the comprehensive behavior score based on the number of normal points and the total number of abnormal points and normal points.

[0029] Preferably, the process of obtaining a recording result based on the comprehensive behavior score and the behavior value, and generating an alarm signal according to the obtained recording result includes:

[0030] Send the face vector without a behavior history line and the monitoring capture data of the abnormal points corresponding to the behavior values of the residents to the patrol personnel to supervise the residents' behaviors, and combine the comprehensive behavior scores of the residents to record the residents' behaviors and obtain a recording result, where the recording result includes a normal recording result and an abnormal recording result; when an abnormal recording result is obtained, generate an alarm signal, and when the alarm signal is received, the patrol personnel implement security measures for the community to obtain a security strategy.

[0031] The present invention also discloses a community intelligent security alarm system based on big data, including a management center, and the management center is communicatively connected to a data acquisition module, a data analysis module, a data processing module, and a security alarm module:

[0032] The data acquisition module is used to obtain access control identification data and monitoring capture data, set up a community database, and construct a community security network according to the community database, access control identification data, and monitoring capture data;

[0033] The data analysis module is used to analyze the access control identification data and monitoring capture data through the community security network to obtain a face vector and a verification group vector, and analyze the face vector and the verification group vector to obtain a tracking value;

[0034] The data processing module is used to obtain a behavior history line according to the community security network and the face vector, and obtain a behavior value and a comprehensive behavior score according to the behavior history line and the tracking value;

[0035] The security alarm module is used to obtain a recording result through the comprehensive behavior score and the behavior value, and generate an alarm signal and a security strategy according to the obtained recording result.

[0036] Compared with the prior art, the above technical solution of the present invention has the following beneficial technical effects:

[0037] By setting up a community database and constructing a community security network, the security response speed and accuracy are improved; potential safety hazards and crime risks are reduced; intelligent community security management is realized; through the face vector and the verification group vector, the security of community information management is improved, and a tracking value is obtained, which helps to trace the residents' behaviors; through the behavior history line, behavior value, and comprehensive behavior score, the integrity and accuracy of residents' behavior analysis are increased; by generating an alarm signal and a security strategy, the modern management level of the community is improved, and the sense of security and satisfaction of residents are enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flowchart of an embodiment proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0039] Embodiment 1, as Figure 1As shown in the figure, a community intelligent security alarm method based on big data proposed by the present invention includes the following steps:

[0040] S1. Obtain access control recognition data and monitoring capture data, set up a community database, and construct a community security network based on the community database, access control recognition data, and monitoring capture data;

[0041] S2. Analyze the access control recognition data and monitoring capture data through the community security network to obtain face vectors and verification group vectors, and analyze the face vectors and verification group vectors to obtain tracking values;

[0042] S3. Obtain a behavior history line based on the community security network and face vectors, and obtain behavior values and behavior comprehensive scores based on the behavior history line and tracking values;

[0043] S4. Obtain a record result through the behavior comprehensive score and behavior value, and generate an alarm signal and a security strategy based on the obtained record result.

[0044] It should be further noted that in the specific implementation process, the process of obtaining access control recognition data and monitoring capture data, setting up a community database, and constructing a community security network based on the community database, access control recognition data, and monitoring capture data is as follows:

[0045] The access control recognition data refers to the data generated when face recognition access control and fingerprint recognition access control verify residents, and the acquisition channels of the access control recognition data are access control devices and access control systems, including residents' residential information, fingerprint data, and face image data; the residential information includes residents' basic information and residential information; the residents' basic information includes basic information such as name, age, and mobile phone number; the residential information includes residents' family address information and address location information;

[0046] The monitoring capture data refers to the relevant data captured by real-time monitoring of cameras in the community, including residents' image data, camera position data, and capture time; the residents' image data refers to the human body image data of residents captured by the cameras;

[0047] Set up a community database; the community database is set up in the system of the access control device;

[0048] The community database refers to the entry and verification of face image data, verification fingerprint data, and valid residential information of valid residents in the community, and the establishment of the residential correlation between the verified face image data and the verified fingerprint data and the valid residential information, including verified face image data, verified fingerprint data, valid residential information, and residential correlation, and the real-time update of the community database, and the deletion of the face image data and fingerprint image data of residents who have moved out of the community or invalid residents;

[0049] Analyze the community database, access control recognition data, and surveillance capture data. Authenticate the residential information, fingerprint data, and face image data in the access control recognition data through the verified face image data, verified fingerprint data, and valid residential information in the community database. The authentication process is as follows: Traverse the valid residential information in the community resident data to obtain the valid residential information that is consistent with the residential information.

[0050] If there is no valid residential information in the community database that is consistent with the residential information, the resident corresponding to this residential information is not a resident of this community and will not be allowed to pass through the access control.

[0051] If there is valid residential information in the community database that is consistent with the residential information, obtain the residential-related relationship of the valid residential information, and through the residential-related relationship, obtain the corresponding verified face image data and verified fingerprint data. Through face verification technology and fingerprint verification technology, compare the verified face image data and verified fingerprint data with the face image data and fingerprint data respectively. If there is an inconsistency between the verified face image data and the face image data or between the verified fingerprint data and the fingerprint data, the resident corresponding to this residential information is not a resident of this community and will not be allowed to pass through the access control.

[0052] If both the verified face image data and the face image data or the verified fingerprint data and the fingerprint data are correspondingly consistent, the resident corresponding to this residential information is a resident of this community, passes through the access control, and records the resident corresponding to this residential information as a monitored resident.

[0053] Set up surveillance data points and access control points. The surveillance data points are set through camera position data; the access control points refer to the access control devices in the community. Combining GIS technology, mark the corresponding surveillance data points and access control points onto the community map through the camera position data, and construct the data interaction relationship between each surveillance data point and the association relationship between each surveillance data point and the access control point. Link each surveillance data point and access control point through the data interaction relationship and association relationship to obtain a surveillance network. Store the capture time and resident image data into the corresponding surveillance data points, and store the verified face image data, verified fingerprint data, and valid residential information corresponding to the monitored residents into the access control points to obtain a community security network.

[0054] It should be further noted that in the specific implementation process, through the community security network, analyze the access control recognition data and surveillance capture data to obtain a face vector and a verification group vector. The process of analyzing the face vector and the verification group vector to obtain a tracking value is as follows:

[0055] Convert the resident image data in the monitoring data points of the community security network and the verified face image data in the access control points into grayscale images. Extract the face regions in the grayscale images through face detection technology, denoted as the monitoring face images and the verified face images. Segment the verified face images, where the segmentation refers to dividing the verified face images based on face features (such as ears, noses, eyes, etc.) to obtain the verified face feature images;

[0056] Set up a face recognition model, which is a deep learning model for inputting face images and outputting vectors through deep learning of the face images;

[0057] Input the monitoring face images and the verified face feature images into the face recognition model to obtain face vectors and verified vectors. Denote all the verified vectors corresponding to the same verified face image as the verified group vectors. Set up a face code, which is a randomly set exclusive face code for each person. Insert the face code into the last item of each verified vector in the verified group vectors, and denote the obtained verified vectors as the verified marked vectors, and denote the verified group vectors as the verified marked group vectors. Through the community security network, send the verified group vectors to each monitoring data point, analyze the verified marked group vectors and the face vectors, and match each item of the verified marked vectors in the verified marked group vectors with the face vectors. The matching specifically means making there be as many consecutive equal items as possible between the verified marked vectors in the verified marked group vectors and the face vectors. Count the number of consecutive equal items, the number of vectors, and the total number of items in the verified marked group vectors between the face vectors and the verified marked group vectors. Obtain the tracking value of the face vectors according to the number of consecutive equal items, the number of vectors, and the total number of items in the verified marked group vectors between the face vectors and the verified marked group vectors;

[0058]

[0059] Where, x is the number of consecutive equal items between the face vectors and the verified marked group vectors; X is the total number of items in the verified marked group vectors; d is the area of the monitoring face image corresponding to the face vectors; D is the area of the verified face image corresponding to the verified group vectors;

[0060] Set up a tracking threshold;

[0061] When the tracking value of the face vectors is greater than or equal to the tracking threshold, obtain the last item of the verified marked vector group and insert the obtained last item into the last item of the face vectors;

[0062] When the tracking value of the face vectors is less than the tracking threshold, match the face vectors with other verified marked vector groups.

[0063] It should be further noted that in the specific implementation process, the process of obtaining the behavior history line based on the community security network and the face vector, and obtaining the behavior value and the comprehensive behavior score based on the behavior history line and the tracking value is as follows:

[0064] Through the community security network, obtain the last item of the face vector of each monitoring data point, denoted as the tracking item. According to the chronological order of capture times, link the monitoring data points corresponding to the face vectors with the same tracking item and the corresponding resident image data to obtain the behavior history line of the resident;

[0065] Obtain the number of resident images in each monitoring data point of the behavior history line, and obtain the behavior value according to the number of resident images;

[0066]

[0067] Where l is the number of resident images in each monitoring data point of the behavior history line; c is the mean value of the number of resident images of the behavior history line; r is the standard deviation of the number of resident images of the behavior history line;

[0068] Set the behavior standard value;

[0069] When the behavior value is greater than or equal to the behavior standard value, the behavior of the resident in the resident image data corresponding to the behavior value is abnormal, and the monitoring data point corresponding to the resident image data is denoted as an abnormal point;

[0070] When the behavior value is less than the behavior standard value, the behavior of the resident in the resident image data corresponding to the behavior value is normal, and the monitoring data point corresponding to the resident image data is denoted as a normal point;

[0071] Calculate the ratio of the number of normal points to the total number of abnormal points and normal points to obtain the comprehensive behavior score.

[0072] It should be further noted that in the specific implementation process, the process of obtaining the recording result based on the comprehensive behavior score and the behavior value, and generating the alarm signal and the security strategy according to the obtained recording result is as follows:

[0073] Send the face vector without a behavior history line and the monitoring capture data of the abnormal points corresponding to the behavior value of the resident to the patrol personnel to supervise the behavior of the resident, and combine the comprehensive behavior score of the resident to record the behavior of the resident to obtain the recording result. The recording result includes the normal recording result and the abnormal recording result; when the abnormal recording result is obtained, generate an alarm signal, and when the alarm signal is received, the patrol personnel implement security measures for the community to obtain the security strategy.

[0074] Embodiment 2. A community intelligent security alarm system based on big data proposed in the present invention is applied to a community intelligent security alarm method based on big data described in Embodiment 1, and specifically includes a management center, which is communicatively connected with a data acquisition module, a data analysis module, a data processing module, and a security alarm module:

[0075] The data acquisition module is used to obtain access control recognition data and monitoring capture data, set up a community database, and construct a community security network according to the community database, access control recognition data, and monitoring capture data;

[0076] The data analysis module is used to analyze the access control recognition data and monitoring capture data through the community security network to obtain face vectors and verification group vectors, and analyze the face vectors and verification group vectors to obtain tracking values;

[0077] The data processing module is used to obtain a behavior history line according to the community security network and face vectors, and obtain a behavior value and a comprehensive behavior score according to the behavior history line and tracking values;

[0078] The security alarm module is used to obtain a record result through the comprehensive behavior score and behavior value, and generate an alarm signal and a security strategy according to the obtained record result.

[0079] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made without departing from the spirit of the present invention within the knowledge scope of those skilled in the art to which the present invention pertains.

Claims

1. A community intelligent security alarm method based on big data, characterized in that, It includes the following steps: S1. Obtain access control recognition data and monitoring capture data, set up a community database, and construct a community security network based on the community database, access control recognition data, and monitoring capture data; S2. Analyze the access control recognition data and monitoring capture data through the community security network to obtain face vectors and verification group vectors, and analyze the face vectors and verification group vectors to obtain tracking values; S3. Obtain a behavior history line based on the community security network and face vectors, and obtain a behavior value and a comprehensive behavior score based on the behavior history line and tracking values; S4. Obtain a record result through the comprehensive behavior score and behavior value, and generate an alarm signal and a security strategy based on the obtained record result.

2. The method for community intelligent security alarm based on big data according to claim 1, wherein The process of obtaining access control recognition data and monitoring capture data, setting up a community database, and constructing a community security network includes: The access control recognition data includes residents' residential information, fingerprint data, and face image data; the monitoring capture data includes residents' image data, camera location data, and capture time; Set up a community database; the community database includes verified face image data, verified fingerprint data, valid residential information, residential related relationships, and a community map; Analyze the community database, access control recognition data, and monitoring capture data to obtain monitored residents; Set up monitoring data points and access control points, combine GIS technology, and construct data interaction relationships and association relationships based on camera location data, monitoring data points, access control points, and the community map. Obtain a monitoring network based on the data interaction relationships, association relationships, monitoring data points, and access control points. Store the capture time and residents' image data in the corresponding monitoring data points, and store the residential information, fingerprint data, and face image data corresponding to the monitored residents in the access control points to obtain a community security network.

3. The community intelligent security alarm method based on big data according to claim 2, characterized in that, The process of analyzing the community database, access control recognition data, and monitoring capture data to obtain monitored residents includes: Obtain valid residential information consistent with the residential information based on the community database and access control recognition data; If there is valid residential information consistent with the residential information in the community database, obtain the residential related relationships of the valid residential information, and obtain the corresponding verified face image data and verified fingerprint data through the residential related relationships. Use face verification technology and fingerprint verification technology to compare the verified face image data and verified fingerprint data with the face image data and fingerprint data respectively. If the verified face image data is corresponding and consistent with the face image data or the verified fingerprint data is corresponding and consistent with the fingerprint data, the resident corresponding to the residential information is a resident of this community, passes through the access control, and record the resident corresponding to the residential information as a monitored resident.

4. The community intelligent security alarm method based on big data according to claim 3, wherein, The process of analyzing the access control recognition data and monitoring capture data through the community security network to obtain face vectors and verification group vectors, and analyzing the face vectors and verification group vectors to obtain tracking values is as follows: Convert the resident image data in the monitoring data points of the community security network and the verified face image data in the access control points into grayscale images. Extract the face regions in the grayscale images through face detection technology to obtain the monitoring face images and the verified face images. Obtain the verified face feature images according to the verified face images. Set up a face recognition model. Input the monitoring face images and the verified face feature images into the face recognition model to obtain face vectors and verified vectors. Obtain the verified group vectors according to the verified face images and the verified vectors. Set up face encoding. Obtain the verified marker vectors and the verified marker group vectors according to the face encoding, the verified group vectors, and the verified vectors. Through the community security network, match each item of the verified marker vectors of the verified marker group vectors with the face vectors. The matching specifically means making the verified marker vectors of the verified marker group vectors have as many consecutive equal items as possible with the face vectors. Count the number of consecutive equal items, the number of vectors, and the total number of items of the verified marker group vectors between the face vectors and the verified marker group vectors. Obtain the tracking value of the face vectors according to the number of consecutive equal items, the number of vectors, and the total number of items of the verified marker group vectors between the face vectors and the verified marker group vectors. Set up a tracking threshold. When the tracking value of the face vectors is greater than or equal to the tracking threshold, obtain the last item of the verified marker vector group and insert the obtained last item into the last item of the face vectors.

5. The community intelligent security alarm method based on big data according to claim 4, characterized in that, The process of obtaining the behavior history line according to the community security network and the face vectors includes: Through the community security network, obtain the last item of the face vectors of each monitoring data point, denoted as the tracking item. Link the monitoring data points corresponding to the face vectors with the same tracking item and the corresponding resident image data in the order of the capture time to obtain the behavior history line of the resident.

6. A community intelligent security alarm method based on big data according to claim 1 or 5, characterized in that, The process of obtaining the behavior value and the behavior comprehensive score according to the behavior history line and the tracking value includes: Obtain the number of resident images in each monitoring data point of the behavior history line and obtain the behavior value according to the number of resident images. Set up a behavior standard value. When the behavior value is greater than or equal to the behavior standard value, mark the monitoring data point corresponding to the resident image data as an abnormal point. When the behavior value is less than the behavior standard value, mark the monitoring data point corresponding to the resident image data as a normal point. Count the number of normal points and the total number of abnormal points and normal points, and obtain the behavior comprehensive score according to the number of normal points and the total number of abnormal points and normal points.

7. The method for community intelligent security alarm based on big data according to claim 6, characterized in that, The process of obtaining the recording result through the behavior comprehensive score and the behavior value and generating an alarm signal according to the obtained recording result includes: Send the face vectors without a behavior history line and the monitoring capture data of the abnormal points corresponding to the behavior values of the residents to the patrol personnel to supervise the residents' behaviors. Combine the behavior comprehensive scores of the residents to record the residents' behaviors and obtain the recording result. The recording result includes a normal recording result and an abnormal recording result. When an abnormal recording result is obtained, generate an alarm signal. When the alarm signal is received, the patrol personnel implement security measures for the community to obtain the security strategy.

8. A community intelligent security alarm system based on big data, specifically applied to a community intelligent security alarm method based on big data as described in any one of claims 1 to 7, including a management center, characterized in that, The management center is communicatively connected to a data acquisition module, a data analysis module, a data processing module, and a security alarm module: The data acquisition module is used to obtain access control identification data and monitoring capture data, set up a community database, and construct a community security network based on the community database, access control identification data, and monitoring capture data; The data analysis module is used to analyze the access control identification data and monitoring capture data through the community security network to obtain face vectors and verification group vectors, and analyze the face vectors and verification group vectors to obtain tracking values; The data processing module is used to obtain a behavior history line based on the community security network and face vectors, and obtain a behavior value and a comprehensive behavior score based on the behavior history line and tracking values; The security alarm module is used to obtain a recording result through the comprehensive behavior score and behavior value, and generate an alarm signal and a security strategy based on the obtained recording result.

Images