Campus security management system based on face recognition

Abstract

The invention discloses a campus security and protection management system based on face recognition, which relates to the technical field of campus security and protection management and comprises a campus security and protection monitoring data acquisition module, a campus security and protection report generation module, a campus security and protection verification module, a campus security and protection management scheme output module and an execution module. Collecting campus security monitoring data including campus user basic data, face data, authority data, access data, campus environment data and positioning data; the campus security verification module analyzes a campus security management report and performs face verification, authority verification and security verification on campus users, and a data acquisition technology, a facial feature extraction technology, a multiple linear regression correction technology and a multi-dimensional verification technology in the system are closely combined with a modern information technology; the problems that traditional campus security and protection management is low in efficiency, easy to omit, difficult in personnel information verification and difficult in tracing are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of campus security management, and particularly relates to a campus security management system based on face recognition. Background Art

[0002] Under the background of the continuous expansion of today's campus scale and the increasing frequency of personnel flow, campus security faces many challenges. Traditional security management methods are gradually difficult to meet the actual needs. Traditional campus security management mainly relies on manual registration of visitor information and manual patrols to ensure campus security. Its efficiency is low and it is prone to omissions. When manually registering visitors, it is difficult to verify the authenticity of the information. Once a security incident occurs,

[0003] the difficulty of tracing is relatively large. At the same time, there are blind spots in time and space for manual patrols, and it is impossible to monitor each campus area in real time. With the rapid development of information technology, face recognition technology has gradually matured and been widely used in many fields. It has advantages such as high accuracy and fast recognition speed, and can effectively make up for the deficiencies of traditional campus security management. Applying face recognition technology to the campus security management system can realize the automated management of personnel access, quickly and accurately identify the identities of campus users, deploy face recognition devices in the campus security management area, record the personnel access situation in real time, and issue corresponding warnings for abnormal situations;

[0004] Although the existing technology has made great progress in the direction of campus security management, there are still some problems to be optimized. The demand for campus security is relatively high. To ensure campus security, means such as visitor registration and security monitoring are usually used for campus security management. However, due to the existing management means, it is difficult to take into account different areas of the entire campus and issue corresponding warnings, which will lead to insufficient security in campus security management. Therefore, how to combine face recognition and security verification to achieve comprehensive campus security management is the problem to be solved by the present invention. For this reason, a campus security management system based on face recognition is proposed. Summary of the Invention

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A campus security management system based on face recognition includes a campus security monitoring data collection module, a campus security report generation module, a campus security verification module, a campus security management plan output module, and an execution module. Among them, each module is communicatively connected;

[0006] The campus security monitoring data collection module collects campus security monitoring data including campus user basic data, face data, permission data, access data, campus environment data, and location data, providing a data basis for subsequent face verification, permission verification, and security verification;

[0007] The campus security report generation module is used to process the pre - processed campus security monitoring data and then generate a campus security report;

[0008] The campus security verification module analyzes the campus security management report, conducts face verification, permission verification and security verification on campus users, and provides a reference basis for outputting the subsequent campus security management plan;

[0009] The campus security management plan output module outputs the corresponding campus security management plan based on the verification results of campus users;

[0010] The execution module takes corresponding security management measures for campus users according to the generated campus security management plan, solving the problem that it is difficult to combine face recognition and security verification in the prior art to achieve comprehensive campus security management.

[0011] A further improvement of the technical solution of the present invention is that: the campus security report generation module is divided into a facial feature extraction unit and a report generation unit. The functions of each unit are as follows:

[0012] The facial feature extraction unit extracts features from the pre - processed face data to obtain the facial feature data of campus users;

[0013] The report generation unit uses the pre - processed environmental data to correct the facial feature data and then generates a campus security management report.

[0014] A further improvement of the technical solution of the present invention is that: the process of the campus security monitoring data acquisition module for acquiring campus security monitoring data includes:

[0015] The campus area is divided into a security area and a non - security area. Among them, the security area includes a functional area and a non - functional area. Combining different types of acquisition devices and data entry technologies, campus security management data of the campus area is acquired;

[0016] The acquisition devices include cameras, light sensors, digital sun trackers, electromagnetic radiation detectors and GPS receivers;

[0017] The basic data of campus users includes the name, ID, gender and identity of campus users. Among them, the identity includes students, faculty and staff, and visitors; the face data is the facial image of campus users; the permission data is the permissions of campus users in the functional area and the non - functional area; the access data is the entry time and departure time of campus users in the campus area; the campus environmental data includes monitoring images, light intensity, light angle, electric field strength and magnetic field strength of the campus area; the positioning data is the longitude and latitude coordinates of the campus area;

[0018] Specifically, the basic data, permission data, and access data of campus users are collected using data entry technology. It should be specifically noted that during the process of collecting permission data, campus users submit reservation access applications for security areas. The management terminal obtains the access permissions of campus users in security areas through data entry. When the reservation access application for a security area is approved, the corresponding campus user has access permission to the security area; when the reservation access application for a security area is not approved, the corresponding campus user does not have access permission to the security area. Facial images of campus users and monitoring images of campus areas are collected using cameras. The light intensity, light angle, electric field intensity, magnetic field intensity, and longitude and latitude coordinates of the campus area are collected through a light sensor, a digital sun tracker, an electromagnetic radiation detector, and a GPS receiver respectively.

[0019] Data cleaning is performed on the collected basic data, permission data, access data, light intensity, light angle, electric field intensity, and magnetic field intensity of the campus area, as well as the positioning data of campus users. Image denoising and image enhancement processing are performed on the facial data and monitoring images of the campus area.

[0020] Timestamps are assigned to the basic data, facial data, permission data, access data, campus environment data, and positioning data of campus users. By adjusting the timestamps, the collection times of the basic data, facial data, permission data, access data, campus environment data, and positioning data of campus users are synchronized. The campus environment data is integrated to generate a campus security monitoring data set. The campus security monitoring data set is divided into a training set and a test set, and the ratio of the training set to the test set is 7:3.

[0021] A further improvement of the technical solution of the present invention lies in: for the facial feature extraction unit, the process of obtaining the facial feature data of campus users includes:

[0022] Using the Haar cascade detector, the face area is detected from the facial images of campus users. Facial key points are detected through the convolutional neural network algorithm. Based on the generative adversarial network method, combining the generator and the discriminator, the coordinates of the facial key points are generated. The facial image key points include key points of both eyes, key points of the nose, key points of the mouth, key points of the eyebrows, key points of the hairline, key points of the left and right temples, key points of the left and right cheekbones, key points of the left and right mandibular angles, key points of the mandible, and key points of the chin tip.

[0023] Using the Euclidean distance formula, the distances between the detected facial image key points are calculated to obtain the facial feature data of campus users. The facial feature data of campus users consists of five - sense organ spacing data and facial contour data. Among them, the five - sense organ spacing data includes the distance between eyes, the distance between eyes and nose, the distance between eyes and mouth, and the distance between nose and mouth. The facial contour data includes the height of the forehead, the width of the forehead, the distance between cheekbones, the length of the mandible, and the width of the mandible. The above - mentioned Euclidean distance formula is as follows:

[0024]

[0025] Among them, (x1, y1) and (x1, x2) are the coordinates of the key points of the facial image to be measured respectively, and d is the distance between two key points of the facial image.

[0026] A further improvement of the technical solution of the present invention lies in: for the report generation unit, the generation process of the campus security management report includes:

[0027] Combining the Haar cascade detector, the convolutional neural network algorithm and the generative adversarial network method, feature extraction is performed on the monitoring images in the campus area to obtain the facial feature data in the monitoring images. It should be noted that the feature extraction principle here is the same as that of the facial feature extraction unit, and the facial feature data in the monitoring images is of the same type as the facial feature data of campus users, and also includes the data of the distances between facial features and the facial contour data. The data of the distances between facial features consists of the distance between eyes, the distance between eyes and nose, the distance between eyes and mouth, and the distance between nose and mouth. The facial contour data consists of the height of the forehead, the width of the forehead, the distance between cheekbones, the length of the lower jaw, and the width of the lower jaw;

[0028] Extract the campus environment data in the campus security monitoring dataset. Using the training set data and combining the multiple linear regression algorithm, the light intensity, light angle, electric field intensity, and magnetic field intensity in the campus area are used as inputs, and the facial feature correction coefficient is used as the output to learn the linear relationship between the light intensity, light angle, electric field intensity, magnetic field intensity in the campus area and the facial feature correction coefficient, and train the facial feature correction model;

[0029] Input the test set data into the facial feature correction model, adjust the intercept term and regression coefficient of the facial feature correction model, optimize the facial feature correction model, deploy the optimized facial feature correction model into the system, and combine the light intensity, light angle, electric field intensity, and magnetic field intensity in the campus area to output the corresponding facial feature correction coefficient;

[0030] The expression of this facial feature correction model is:

[0031] U = α0 + α1u1 + α2u2 + α3u3 + α4u4 + θ

[0032] Among them, U is the facial feature correction coefficient; α1, α2, α3, and α4 are the regression coefficients of the light intensity, light angle, electric field intensity, and magnetic field intensity in the campus area respectively; u1, u2, u3, and u4 are the light intensity, light angle, electric field intensity, and magnetic field intensity in the campus area respectively; α0 and θ are the intercept term and regression coefficient of the facial feature correction model respectively;

[0033] Using the facial feature correction coefficient, correct the facial feature data in the surveillance image to obtain the security identification facial feature data. Specifically, for the facial feature data in the surveillance image, when the facial feature correction coefficient is lower than 0.1, do not correct the facial feature data of campus users in the surveillance image; when the facial feature correction coefficient is between 0.1 and 0.2, between 0.2 and 0.3, between 0.3 and 0.4, between 0.4 and 0.5, between 0.5 and 0.6, between 0.6 and 0.7, between 0.7 and 0.8, between 0.8 and 0.9, and between 0.9 and 1, correct the interpupillary distance, the distance between the eyes and the nose, the distance between the eyes and the mouth, the distance between the nose and the mouth, the forehead height, the forehead width, the zygomatic distance, the jaw length, and the jaw width respectively;

[0034] Integrate the preprocessed basic data, permission data, access data, positioning data, facial feature data of campus users, and security identification facial feature data of campus users to generate a campus security management report.

[0035] A further improvement of the technical solution of the present invention lies in: the process of the campus security verification module for face verification of campus users includes:

[0036] Through the campus security management report, compare the facial feature data of campus users with the security identification facial feature data. If the facial feature data of campus users exceeds 0% to 5% of the corresponding security identification facial feature data, the face verification of campus users passes; if the facial feature data of campus users exceeds 5% of the corresponding security identification facial feature data, the face verification of campus users fails.

[0037] A further improvement of the technical solution of the present invention lies in: the process of the campus security verification module for permission verification of campus users includes:

[0038] According to the face verification result, when the face verification of campus users fails, no permission verification is performed; when the face verification of campus users passes, permission verification is performed;

[0039] Set that campus users have permission to access non-security areas, and only campus users who have passed the permission verification have permission to access security areas;

[0040] According to the identities of campus users, access permissions for security areas are assigned to campus users. The security areas include functional areas and non-functional areas. When the campus users are students and faculty members, they have the right to access non-functional areas and do not need permission verification to access functional areas. Permission verification for functional areas is required for campus users with the identities of students and faculty members. When the campus user is a visitor, they do not have permission to access functional areas and non-functional areas, and permission verification for the entire functional area and non-functional area is required for campus users with the identity of visitors. Specifically, non-security areas are areas such as campus roads, canteens, and public toilets; security areas include functional areas and non-functional areas. Among them, non-functional areas include areas such as libraries, dormitories, teaching buildings, and playgrounds; functional areas include areas such as laboratories and power distribution rooms;

[0041] Based on the permission data in the campus security management report, analyze the permissions of campus users in functional areas and non-functional areas. When the access permission of a campus user to the corresponding security area fails, the campus user is not allowed to access the corresponding security area; when the access permission of a campus user to the corresponding security area passes, the campus user is allowed to access the corresponding security area.

[0042] A further improvement of the technical solution of the present invention lies in: the process of the campus security verification module for security verification of campus users includes:

[0043] After the access permission of a campus user in the security area passes, security verification is performed on the campus user. Through the access and exit data in the campus security management report, analyze the entry time and exit time of the campus user in the campus area, calculate the access and exit time difference of the campus user in the campus area, and count the number of access and exit times of the campus user in the campus area within 24 hours;

[0044] Based on the access and exit time difference of the campus user in the campus area and the number of access and exit times within 24 hours, the first security coefficient and the second security coefficient are respectively assigned to the campus user;

[0045] Specifically, when the access and exit time difference of a campus user in the campus area is less than 10 hours, the first security coefficient assigned to the corresponding campus user is 0.3; when the access and exit time difference of a campus user in the campus area is between 10 hours and 18 hours, the first security coefficient assigned to the corresponding campus user is 0.5; when the access and exit time difference of a campus user in the campus area exceeds 18 hours, the first security coefficient assigned to the corresponding campus user is 0.8; similarly, when the number of access and exit times of a campus user in the campus area within 24 hours is less than 5 times, the second security coefficient assigned to the corresponding campus user is 0.3; when the number of access and exit times of a campus user in the campus area within 24 hours is between 5 times and 8 times, the second security coefficient assigned to the corresponding campus user is 0.6; when the number of access and exit times of a campus user in the campus area within 24 hours is more than 5 times, the second security coefficient assigned to the corresponding campus user is 0.9;

[0046] Weights are assigned to the first safety factor and the second safety factor respectively. Through the weight summation method, the safety verification coefficient of campus users is calculated. The calculation process includes:

[0047] H = w1 × h1 + w2 × h2

[0048] Where H is the safety verification coefficient of campus users, w1 and w2 are the weights of the first safety factor and the second safety factor respectively, and h1 and h2 are the first safety factor and the second safety factor respectively;

[0049] Based on the calculated safety verification coefficient of campus users, the safety verification of campus users is carried out. When the safety verification coefficient of campus users is lower than 0.4, it indicates that the corresponding campus user's safety verification is passed; when the safety verification coefficient of campus users is greater than 0.4, it indicates that the corresponding campus user's safety verification fails.

[0050] A further improvement of the technical solution of the present invention lies in: for the campus security management plan output module, the output process of the campus security management plan includes:

[0051] Based on the face verification result, when the face verification result is not passed, an alarm signal for an abnormal person entering the campus area is issued to remind the security personnel to intervene and handle it; when the face verification result is passed, no alarm signal is issued;

[0052] Based on the permission verification result, the permission verification of the security area is carried out. When the permission verification of the campus user fails, the corresponding campus user is not allowed to enter the security area; when the permission verification is passed, the corresponding campus user is allowed to enter the security area;

[0053] Based on the safety verification result, when the safety verification of the campus user fails, a campus user safety alarm signal is issued, and when the safety verification of the campus user is passed, no campus safety alarm signal is issued, and then the campus security management plan is output.

[0054] A further improvement of the technical solution of the present invention lies in: for the execution module, the process of taking corresponding security management measures for campus users according to the generated campus security management plan includes:

[0055] Combined with the actual face verification result, permission verification result and safety verification result of campus users, according to the generated campus security management plan, corresponding alarm signals are issued to notify the security personnel to take corresponding measures to manage the campus security.

[0056] The beneficial effects of the present invention are as follows: In the campus security management system based on face recognition of the present invention, compared with the traditional campus security management system based on face recognition, the data acquisition technology, facial feature extraction technology, multiple linear regression correction technology, multi-dimensional verification technology in the system of the present invention are closely combined with modern information technology to accurately capture the basic data of campus users, face data, permission data, access data, campus environment data and positioning data, obtain and correct the detailed facial feature data of campus users, and then achieve face verification, permission verification and security verification of campus users, solve the problems of low efficiency, easy omission, difficult verification of personnel information and difficult traceability in traditional campus security management, ensure that the method in the present invention can refine the dynamic monitoring standard for a campus security management system based on face recognition within a more accurate range, make the monitored data become more accurate indicators under the same conditions. The research and application of this method significantly enhance the degree of intelligence in the campus security management process based on face recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0058] Figure 1 It is a block diagram of a campus security management system based on face recognition of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0059] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0060] As Figure 1 shown, the present invention provides a campus security management system based on face recognition, including a campus security monitoring data acquisition module, a campus security report generation module, a campus security verification module, a campus security management plan output module and an execution module. Among them, each module is communicatively connected;

[0061] The campus security monitoring data acquisition module collects campus security monitoring data including basic data of campus users, face data, permission data, access data, campus environment data and positioning data, providing a data basis for subsequent face verification, permission verification and security verification;

[0062] The campus security report generation module is used to process the pre - processed campus security monitoring data and then generate a campus security report;

[0063] The campus security verification module analyzes the campus security management report, conducts face verification, permission verification and security verification on campus users, providing a reference basis for outputting the subsequent campus security management plan;

[0064] The campus security management plan output module outputs the corresponding campus security management plan based on the verification results of campus users;

[0065] The execution module takes corresponding security management measures for campus users according to the generated campus security management plan, solving the problem that it is difficult to combine face recognition and security verification in the prior art to achieve comprehensive campus security management.

[0066] The campus security report generation module is divided into a facial feature extraction unit and a report generation unit. The functions of each unit are as follows:

[0067] The facial feature extraction unit extracts features from the pre - processed face data to obtain the facial feature data of campus users;

[0068] The report generation unit uses the pre - processed environmental data to correct the facial feature data and then generates a campus security management report.

[0069] The process of the campus security monitoring data acquisition module for collecting campus security monitoring data includes:

[0070] Dividing the campus area into security areas and non - security areas. Among them, the security areas include functional areas and non - functional areas, and combining different types of acquisition devices and data entry technologies to collect campus security management data in the campus area;

[0071] The acquisition devices include cameras, light sensors, digital sun trackers, electromagnetic radiation detectors and GPS receivers;

[0072] The basic data of campus users includes the name, ID, gender and identity of campus users. Among them, the identity includes students, faculty and staff, and visitors; the face data is the facial image of campus users; the permission data is the permissions of campus users in functional areas and non - functional areas; the access data is the entry time and departure time of campus users in the campus area; the campus environmental data includes the monitoring images, light intensity, light angle, electric field intensity and magnetic field intensity of the campus area; the positioning data is the longitude and latitude coordinates of the campus area;

[0073] Specifically, data entry technology is used to collect basic campus user data, permission data, and access data. It should be specifically noted that during the process of collecting permission data, campus users submit reservation access applications for security areas. The management terminal obtains the access permissions of campus users in security areas through data entry. When the reservation access application for a security area is approved, the corresponding campus user has access permission to the security area. When the reservation access application for a security area is not approved, the corresponding campus user does not have access permission to the security area; cameras are used to collect facial images of campus users and surveillance images of campus areas. The light intensity, light angle, electric field intensity, magnetic field intensity, and longitude and latitude coordinates of the campus area are collected through a light sensor, a digital sun tracker, an electromagnetic radiation detector, and a GPS receiver respectively;

[0074] Data cleaning is performed on the collected basic campus user data, permission data, access data, light intensity, light angle, electric field intensity, and magnetic field intensity of the campus area, as well as positioning data. Image denoising and image enhancement processing are performed on the face data and surveillance images of the campus area;

[0075] Timestamps are assigned to the basic campus user data, face data, permission data, access data, campus environment data, and positioning data. By adjusting the timestamps, the collection times of the basic campus user data, face data, permission data, access data, campus environment data, and positioning data are synchronized. The campus environment data is integrated to generate a campus security monitoring dataset, and the campus security monitoring dataset is divided into a training set and a test set, and the ratio of the training set to the test set is 7:3.

[0076] Facial feature extraction unit. The process of obtaining facial feature data of campus users includes:

[0077] Using the Haar cascade detector, the face area is detected from the facial images of campus users. Facial key points are detected through the convolutional neural network algorithm. Based on the generative adversarial network method, combining the generator and the discriminator, the coordinates of the facial key points are generated. The facial image key points include key points of both eyes, nose key points, mouth key points, eyebrow key points, hairline key points, left and right temple key points, left and right cheekbone key points, left and right mandibular angle key points, mandible key points, and chin tip key points.

[0078] Using the Euclidean distance formula, the distances between the detected facial image key points are calculated to obtain the facial feature data of campus users. The facial feature data of campus users consists of five sense organ spacing data and facial contour data. Among them, the five sense organ spacing data includes eye spacing, eye-nose spacing, eye-mouth spacing, and nose-mouth spacing. The facial contour data includes forehead height, forehead width, cheekbone spacing, mandible length, and mandible width. The above Euclidean distance formula is as follows:

[0079]

[0080] Among them, (x1, y1) and (x1, x2) are the coordinates of the key points of the facial image to be measured, and d is the distance between two key points of the facial image.

[0081] A report generation unit, and the generation process of the campus security management report includes:

[0082] Combining the Haar cascade detector, the convolutional neural network algorithm and the generative adversarial network method, extracting features from the monitoring images in the campus area to obtain the facial feature data in the monitoring images. It should be noted that the feature extraction principle here is the same as that of the facial feature extraction unit, and the facial feature data in the monitoring images is of the same type as the facial feature data of campus users, and also includes the data of the distances between facial features and the facial contour data. The data of the distances between facial features consists of the distance between eyes, the distance between eyes and nose, the distance between eyes and mouth, and the distance between nose and mouth. The facial contour data consists of the forehead height, forehead width, zygomatic distance, jaw length and jaw width;

[0083] Extracting the campus environment data from the campus security monitoring dataset, using the training set data, and combining the multiple linear regression algorithm, taking the light intensity, light angle, electric field intensity and magnetic field intensity in the campus area as inputs and the facial feature correction coefficient as the output, learning the linear relationship between the light intensity, light angle, electric field intensity, magnetic field intensity in the campus area and the facial feature correction coefficient, and training the facial feature correction model;

[0084] Inputting the test set data into the facial feature correction model, adjusting the intercept term and regression coefficient of the facial feature correction model, optimizing the facial feature correction model, deploying the optimized facial feature correction model into the system, and combining the light intensity, light angle, electric field intensity and magnetic field intensity in the campus area to output the corresponding facial feature correction coefficient;

[0085] The expression of this facial feature correction model is:

[0086] U = α0 + α1u1 + α2u2 + α3u3 + α4u4 + θ

[0087] Among them, U is the facial feature correction coefficient; α1, α2, α3 and α4 are the regression coefficients of the light intensity, light angle, electric field intensity and magnetic field intensity in the campus area respectively; u1, u2, u3 and u4 are the light intensity, light angle, electric field intensity and magnetic field intensity in the campus area respectively; α0 and θ are the intercept term and regression coefficient of the facial feature correction model respectively;

[0088] Using the facial feature correction coefficient, correct the facial feature data in the surveillance image to obtain the security identification facial feature data. Specifically, for the facial feature data in the surveillance image, when the facial feature correction coefficient is lower than 0.1, do not correct the facial feature data of campus users in the surveillance image; when the facial feature correction coefficient is between 0.1 and 0.2, between 0.2 and 0.3, between 0.3 and 0.4, between 0.4 and 0.5, between 0.5 and 0.6, between 0.6 and 0.7, between 0.7 and 0.8, between 0.8 and 0.9, and between 0.9 and 1, correct the interpupillary distance, the distance between the eyes and the nose, the distance between the eyes and the mouth, the distance between the nose and the mouth, the forehead height, the forehead width, the zygomatic distance, the mandible length, and the mandible width respectively;

[0089] Integrate the preprocessed basic data, permission data, access data, location data, facial feature data of campus users, and security identification facial feature data of campus users to generate a campus security management report.

[0090] The process of the campus security verification module for face verification of campus users includes:

[0091] Through the campus security management report, compare the facial feature data of campus users with the security identification facial feature data. If the facial feature data of campus users exceed 0% - 5% of the corresponding security identification facial feature data, the face verification of campus users passes; if the facial feature data of campus users exceed 5% of the corresponding security identification facial feature data, the face verification of campus users fails.

[0092] The process of the campus security verification module for permission verification of campus users includes:

[0093] According to the face verification result, when the face verification of campus users fails, no permission verification is performed; when the face verification of campus users passes, permission verification is performed;

[0094] Set that campus users have permission to access non-security areas, and only campus users who pass the permission verification have permission to access security areas;

[0095] According to the identities of campus users, access permissions for security areas are assigned to campus users. The security areas include functional areas and non-functional areas. When the campus users are students and teaching staff, they have the right to access non-functional areas, but there is no permission to verify access to functional areas. It is necessary to verify the permissions of campus users with the identities of students and teaching staff for functional areas. When the campus users are visitors, they have no permission to access functional areas and non-functional areas. It is necessary to verify the permissions of campus users with the identity of visitors for the entire functional area and non-functional area. Specifically, the non-security areas are areas such as campus roads, canteens, and public toilets; the security areas include functional areas and non-functional areas. Among them, the non-functional areas include areas such as libraries, dormitories, teaching buildings, and playgrounds; the functional areas include areas such as laboratories and power distribution rooms;

[0096] Based on the permission data in the campus security management report, analyze the permissions of campus users in functional areas and non-functional areas. When the access permission of a campus user to the corresponding security area fails, the campus user is not allowed to access the corresponding security area; when the access permission of a campus user to the corresponding security area passes, the campus user is allowed to access the corresponding security area.

[0097] The campus security verification module. The process of security verification for campus users includes:

[0098] After the access permission of a campus user in the security area passes, conduct a security verification on the campus user. Analyze the entry time and departure time of the campus user in the campus area through the access data in the campus security management report, calculate the time difference between the entry and departure of the campus user in the campus area, and count the number of times the campus user enters and exits the campus area within 24 hours;

[0099] Based on the time difference between the entry and departure of the campus user in the campus area and the number of times the campus user enters and exits within 24 hours, assign a first security coefficient and a second security coefficient to the campus user respectively;

[0100] Specifically, when the time difference between the entry and departure of a campus user in the campus area is less than 10 hours, the first security coefficient assigned to the corresponding campus user is 0.3; when the time difference between the entry and departure of a campus user in the campus area is between 10 hours and 18 hours, the first security coefficient assigned to the corresponding campus user is 0.5; when the time difference between the entry and departure of a campus user in the campus area exceeds 18 hours, the first security coefficient assigned to the corresponding campus user is 0.8; similarly, when the number of times a campus user enters and exits the campus area within 24 hours is less than 5 times, the second security coefficient assigned to the corresponding campus user is 0.3; when the number of times a campus user enters and exits the campus area within 24 hours is between 5 times and 8 times, the second security coefficient assigned to the corresponding campus user is 0.6; when the number of times a campus user enters and exits the campus area within 24 hours is more than 5 times, the second security coefficient assigned to the corresponding campus user is 0.9;

[0101] Weights are assigned to the first safety factor and the second safety factor for distribution. Through the method of weight summation, the safety verification coefficient of campus users is calculated. The calculation process includes:

[0102] H = w1×h1 + w2×h2

[0103] Where H is the safety verification coefficient of campus users, w1 and w2 are the weights of the first safety factor and the second safety factor respectively, and h1 and h2 are the first safety factor and the second safety factor respectively;

[0104] Based on the calculated safety verification coefficient of campus users, safety verification of campus users is carried out. When the safety verification coefficient of campus users is lower than 0.4, it indicates that the corresponding campus user's safety verification is passed; when the safety verification coefficient of campus users is greater than 0.4, it indicates that the corresponding campus user's safety verification fails.

[0105] Campus security management plan output module. The output process of the campus security management plan includes:

[0106] Based on the face verification result, when the face verification result is not passed, an alarm signal for abnormal personnel entering the campus area is issued to remind the security personnel to intervene and handle it; when the face verification result is passed, no alarm signal is issued;

[0107] Based on the permission verification result, permission verification is carried out for the security area. When the permission verification of the campus user fails, the corresponding campus user is not allowed to enter the security area; when the permission verification is passed, the corresponding campus user is allowed to enter the security area;

[0108] Based on the safety verification result, when the safety verification of the campus user fails, a campus user safety alarm signal is issued. When the safety verification of the campus user is passed, no campus safety alarm signal is issued, and then the campus security management plan is output.

[0109] Execution module. The process of taking corresponding security management measures for campus users according to the generated campus security management plan includes:

[0110] Combined with the actual face verification results, authority verification results and security verification results of campus users, according to the generated campus security management plan, the corresponding alarm signal is issued to notify the security personnel to take corresponding measures to manage campus security. First, through different types of acquisition equipment and data entry technology, the basic data, face data, authority data, access data, campus environment data and positioning data of campus users are collected; secondly, the pre-processed face data and the monitoring images of the campus area are extracted to obtain the facial feature data of campus users and the facial feature data in the monitoring images. Combined with the pre-processed environmental data and the multivariate linear regression algorithm, the facial feature correction coefficient is output, and then the facial feature data in the monitoring image is corrected to obtain the security recognition facial feature data and generate a campus security management report; then, the campus security management report is analyzed, and the face verification, authority verification and security verification are performed on the campus users; then, based on the verification results of the campus users, the corresponding campus security management plan is output; finally, according to the generated campus security management plan, corresponding security management measures are taken for campus users.

[0111] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A campus security management system based on face recognition, including a campus security monitoring data collection module, a campus security report generation module, a campus security verification module, a campus security management plan output module, and an execution module. Among them, Each module is communicatively connected, and is characterized in that: The campus security monitoring data acquisition module acquires campus security monitoring data including campus user basic data, face data, permission data, access data, campus environment data, and positioning data; The campus security report generation module is used to process the preprocessed campus security monitoring data, and then generate a campus security report; The campus security verification module analyzes the campus security management report and conducts face verification, permission verification, and security verification on campus users; The campus security management solution output module outputs corresponding campus security management solutions based on the verification results of campus users; The execution module takes corresponding security management measures for campus users according to the generated campus security management solutions.

2. The campus security management system based on face recognition according to claim 1, wherein: The campus security report generation module is divided into a facial feature extraction unit and a report generation unit. Among them, the functions of each unit are as follows: The facial feature extraction unit extracts features from the preprocessed face data to obtain the facial feature data of campus users; The report generation unit uses the preprocessed environment data to correct the facial feature data, and then generates a campus security management report.

3. The campus security management system based on face recognition according to claim 2, wherein: The process of the campus security monitoring data acquisition module acquiring campus security monitoring data includes: The campus area is divided into security areas and non-security areas. Among them, the security areas include functional areas and non-functional areas. Combining different types of acquisition devices and data entry technologies, campus security management data of the campus area is acquired; The acquisition devices include cameras, light sensors, digital sun trackers, electromagnetic radiation detectors, and GPS receivers; The campus user basic data includes the name, ID, gender, and identity of campus users. Among them, the identity includes students, faculty, and visitors; the face data is the facial image of campus users; the permission data is the permissions of campus users in functional areas and non-functional areas; the access data is the entry time and departure time of campus users in the campus area; the campus environment data includes surveillance images, light intensity, light angle, electric field intensity, and magnetic field intensity of the campus area; the positioning data is the longitude and latitude coordinates of the campus area; Data cleaning is performed on the acquired campus user basic data, permission data, access data, light intensity, light angle, electric field intensity, and magnetic field intensity of the campus area, and positioning data, and image denoising processing and image enhancement processing are performed on the face data and surveillance images of the campus area; Timestamps are assigned to the campus user basic data, face data, permission data, access data, campus environment data, and positioning data. By adjusting the timestamps, the acquisition times of the campus user basic data, face data, permission data, access data, campus environment data, and positioning data are synchronized. The campus environment data is integrated to generate a campus security monitoring data set, and the campus security monitoring data set is divided into a training set and a test set.

4. The campus security management system based on face recognition according to claim 3, characterized in that: The process of the facial feature extraction unit obtaining the facial feature data of campus users includes: Using the Haar cascade detector, the face region is detected from the facial images of campus users. Through the convolutional neural network algorithm, facial key points are detected. Based on the generative adversarial network method, combining the generator and the discriminator, the coordinates of the facial key points are generated; Using the Euclidean distance formula, the distances between the key points of the detected facial images are calculated to obtain the facial feature data of campus users. The facial feature data of campus users consists of five-feature-spacing data and facial contour data. Among them, the five-feature-spacing data includes the distance between eyes, the distance between eyes and nose, the distance between eyes and mouth, and the distance between nose and mouth. The facial contour data includes forehead height, forehead width, zygomatic bone spacing, jaw length, and jaw width.

5. The campus security management system based on face recognition according to claim 4, characterized in that: For the report generation unit, the generation process of the campus security management report includes: Combining the Haar cascade detector, the convolutional neural network algorithm, and the generative adversarial network method, feature extraction is performed on the surveillance images of the campus area to obtain the facial feature data in the surveillance images; The campus environmental data in the campus security monitoring dataset is extracted. Using the training set data and combining the multiple linear regression algorithm, the light intensity, light angle, electric field intensity, and magnetic field intensity in the campus area are used as inputs, and the facial feature correction coefficient is used as the output to learn the linear relationship between the light intensity, light angle, electric field intensity, magnetic field intensity in the campus area and the facial feature correction coefficient, and train the facial feature correction model; The test set data is input into the facial feature correction model, the intercept term and regression coefficient of the facial feature correction model are adjusted, the facial feature correction model is optimized, and the optimized facial feature correction model is deployed into the system. Combining the light intensity, light angle, electric field intensity, and magnetic field intensity in the campus area, the corresponding facial feature correction coefficients are output; Using the facial feature correction coefficient, the facial feature data in the surveillance images is corrected to obtain the security identification facial feature data; Integrate the preprocessed basic data, permission data, access data, location data of campus users, the facial feature data of campus users, and the security identification facial feature data to generate a campus security management report.

6. The campus security management system based on face recognition according to claim 5, characterized in that: For the campus security verification module, the process of face verification of campus users includes: Through the campus security management report, compare the facial feature data of campus users with the security identification facial feature data. If the facial feature data of campus users exceeds 0% - 5% of the corresponding security identification facial feature data, the face verification of campus users passes; if the facial feature data of campus users exceeds 5% of the corresponding security identification facial feature data, the face verification of campus users fails.

7. The campus security management system based on face recognition according to claim 6, characterized in that: For the campus security verification module, the process of permission verification of campus users includes: According to the face verification result, when the face verification of campus users fails, no permission verification is performed; when the face verification of campus users passes, permission verification is performed; It is set that campus users have permission to access non-security areas, and only campus users who have passed the permission verification have permission to access security areas; According to the identities of campus users, access permissions for security areas are assigned to campus users. The security areas include functional areas and non-functional areas. When campus users are students and faculty members, they have permission to access non-functional areas, but do not have permission to access functional areas without authentication. Permission verification for functional areas is required for campus users with the identities of students and faculty members. When campus users are visitors, they do not have permission to access functional areas and non-functional areas, and permission verification for the entire functional area and non-functional area is required for campus users with the identity of visitors. Based on the permission data in the campus security management report, the permissions of campus users in functional areas and non-functional areas are analyzed. When the access permission of a campus user to the corresponding security area fails, access to the corresponding security area is not allowed. When the access permission of a campus user to the corresponding security area passes, access to the corresponding security area is allowed.

8. The campus security management system based on face recognition according to claim 7, characterized in that: In the campus security verification module, the process of security verification for campus users includes: After the access permission of a campus user in the security area passes, security verification is performed on the campus user. Based on the access and exit data in the campus security management report, the entry time and exit time of the campus user in the campus area are analyzed, the difference in access and exit time of the campus user in the campus area is calculated, and the number of access and exit times of the campus user in the campus area within 24 hours is counted. Based on the difference in access and exit time of the campus user in the campus area and the number of access and exit times within 24 hours, a first security coefficient and a second security coefficient are respectively assigned to the campus user. Weights are respectively assigned to the assigned first security coefficient and second security coefficient, and the security verification coefficient of the campus user is calculated through the weight summation method. Based on the calculated security verification coefficient of the campus user, security verification is performed on the campus user. When the security verification coefficient of the campus user is lower than 0.4, it indicates that the security verification of the corresponding campus user passes. When the security verification coefficient of the campus user is greater than 0.4, it indicates that the security verification of the corresponding campus user fails.

9. The campus security management system based on face recognition according to claim 8, characterized in that: In the campus security management plan output module, the output process of the campus security management plan includes: Based on the face verification result, when the face verification result fails, an alarm signal for an abnormal person entering the campus area is issued to remind the security personnel to intervene and handle. When the face verification result passes, no alarm signal is issued. Based on the permission verification result, permission verification for the security area is performed. When the permission verification of a campus user fails, the campus user is not allowed to enter the corresponding security area. After the permission verification passes, the corresponding campus user is allowed to enter the security area. Based on the security verification result, when the security verification of a campus user fails, a security alarm signal for the campus user is issued. When the security verification of the campus user passes, no campus security alarm signal is issued, and then the campus security management plan is output.

10. A campus security management system based on face recognition according to claim 9, characterized in that: In the execution module, the process of taking corresponding security management measures for campus users according to the generated campus security management plan includes: Combining the actual face verification result, permission verification result, and security verification result of the campus user, and based on the generated campus security management plan, corresponding alarm signals are issued to notify the security personnel to take corresponding measures to manage the campus security.

Images