AI-based behavior analysis and recognition system

By introducing AI technology into the campus monitoring system, analyzing video surveillance data to identify abnormal behaviors and triggering early warnings, the defects of existing systems that cannot promptly detect and deal with safety issues are solved, and the active warning and efficient management of campus safety are achieved, and the overall improvement is achieved. Safety level.

CN119027848AInactive Publication Date: 2024-11-26HUIZHI DAOSHENG (WENZHOU) SYST TECH CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410762755.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-13
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing campus monitoring system lacks the functions of analysis, identification and active early warning, and cannot detect and deal with safety issues in a timely manner, resulting in a single and passive security prevention method.

Method used

A behavioral analysis and identification system based on AI was designed to obtain video surveillance data through the data input module, and use the AI ​​analysis model to perform intelligent big data analysis to identify abnormal behaviors and trigger early warnings, including behavioral abnormalities, trajectory abnormalities, abnormal people's number and abnormal list , and push it through the alarm module.

Benefits of technology

It has achieved active early warning and efficient handling on campus, improved the intelligence level of campus safety management, reduced personnel negligence, reduced potential risks, and achieved comprehensive safety monitoring.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119027848A_ABST
    Figure CN119027848A_ABST
Patent Text Reader

Abstract

The embodiment of the invention discloses an AI-based behavior analysis and recognition system, and the system comprises a data input module which is used for obtaining video monitoring data, collected by a video monitoring device, of each point location area of a smart campus; the behavior analysis module is used for sending the obtained video monitoring data into a pre-deployed AI analysis model to perform intelligent big data analysis so as to obtain a corresponding analysis result; the early warning module is used for triggering and generating corresponding early warning data when the analysis result contains at least one preset abnormal behavior; the abnormal behaviors comprise behavior abnormity, track abnormity, people number abnormity and list abnormity; the alarm module is used for carrying out early warning pushing on the generated early warning data; the beneficial effects of the invention are that the AI technology is enabled to be applied to the monitoring system, so that the system can timely find related abnormal behaviors existing in the campus, and initiative early warning and pushing are carried out; therefore, the intelligent level of campus safety management is improved, and the campus safety is changed from passive to active.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of smart campus, and in particular to an AI-based behavior analysis and recognition system. Background Art

[0002] At present, campus security measures mainly rely on a large number of surveillance cameras, combined with facial devices at entrances and exits and key areas to capture tracks, and then aggregate relevant images to the monitoring center; however, the existing monitoring system does not have analysis and recognition and active early warning functions, and can only be used for post-event evidence collection, with a single means of protection; at the same time, for security issues that arise, unless students actively report them after seeing them, they will not be able to discover the problems in time. Summary of the invention

[0003] In view of the technical defects mentioned in the background technology, an embodiment of the present invention aims to provide an AI-based behavior analysis and recognition system.

[0004] To achieve the above objectives, an embodiment of the present invention provides an AI-based behavior analysis and recognition system, the system comprising:

[0005] The data input module is used to obtain the video surveillance data of each point area of ​​the smart campus collected by the video surveillance equipment;

[0006] A behavior analysis module is used to send the obtained video surveillance data to a pre-deployed AI analysis model to perform intelligent big data analysis to obtain corresponding analysis results; wherein the AI ​​analysis model includes multiple recognition sub-models of different scene types;

[0007] An early warning module is used to trigger the generation of corresponding early warning data when the analysis result contains at least one preset abnormal behavior; wherein the abnormal behavior includes abnormal behavior, abnormal trajectory, abnormal number of people and abnormal list;

[0008] The alarm module is used to push the generated warning data.

[0009] As a specific implementation method of the present application, when the warning is triggered, a target person portrait that triggers the warning is immediately captured.

[0010] As a preferred implementation method of the present application, it also includes a trajectory management module, which is used to generate a corresponding trajectory file for each student based on the structured data generated by each point area, combined with the time sequence, location and video surveillance data.

[0011] As a preferred implementation of the present application, the trajectory management module is also used for trajectory control, and the trajectory control includes personal control, group control, regional control and time period control;

[0012] Among them, if there is control, once the student appears in the controlled area, the status is abnormal; if there is no control, the default status is normal.

[0013] As a preferred implementation method of the present application, the trajectory management module is also used for target positioning. When the current location of a student in school needs to be located in real time, the target positioning is used to retrieve the most recent snapshot address by searching by image or name to locate the current location of the student.

[0014] As a preferred implementation of the present application, it also includes an association module, which is used to associate user information, associate personnel information, associate device information, associate terminal information and associate map information.

[0015] As a specific implementation method of the present application, the early warning module is also used for early warning management, which includes linkage early warning, early warning pop-up screen, message handling, alarm reporting, alarm timeout and multiple alarm warnings.

[0016] As a specific implementation method of the present application, the behavioral anomaly is obtained through a deep learning algorithm based on a neural network to obtain the human skeleton of each student in the video screen, and according to the motion trajectory of the student's limb movements, the coordinate changes of the skeleton in the screen are simulated, and then the student behavior is calculated based on the spatiotemporal characteristics of the movement change law.

[0017] As a specific implementation method of the present application, the abnormal number of people is judged by combining the number count of the facial access control channel gate, the entry and exit statistics of the dormitory entrance and exit cameras, the roll call statistics of the cameras in the classroom, early warning data and facial recognition technology.

[0018] The technical solution provided by the embodiment of the present invention obtains video surveillance data collected by video surveillance equipment; sends the obtained video surveillance data to a pre-deployed AI analysis model to perform intelligent big data analysis to obtain corresponding analysis results; when the analysis results contain at least one preset abnormal behavior, triggers the generation of corresponding early warning data and pushes it; by empowering the monitoring system with AI technology, the system can timely discover relevant abnormal behaviors on campus through big data analysis, and actively warn and push; thereby improving the intelligence level of campus safety management, changing campus safety from passive to active, reducing personnel negligence, and through timely and efficient processing, also reducing possible potential risks, thereby achieving the purpose of all-round security monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the specific implementation of the present invention or the technical solution in the prior art, the drawings required for use in the specific implementation or the description of the prior art are briefly introduced below.

[0020] Figure 1 It is a structural diagram of an AI-based behavior analysis and recognition system provided by an embodiment of the present invention;

[0021] Figure 2 It is a schematic diagram of a system architecture provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0024] Please refer to Figure 1 to Figure 2 , an embodiment of the present invention provides an AI-based behavior analysis and recognition system, the system comprising:

[0025] The data input module is used to obtain the video surveillance data of each point area of ​​the smart campus collected by the video surveillance equipment;

[0026] A behavior analysis module is used to send the obtained video surveillance data to a pre-deployed AI analysis model to perform intelligent big data analysis to obtain corresponding analysis results; wherein the AI ​​analysis model includes multiple recognition sub-models of different scene types;

[0027] An early warning module is used to trigger the generation of corresponding early warning data when the analysis result contains at least one preset abnormal behavior; wherein the abnormal behavior includes abnormal behavior, abnormal trajectory, abnormal number of people and abnormal list;

[0028] The alarm module is used to push the generated warning data.

[0029] In this embodiment, the content is pushed to corresponding terminals, including mobile terminals, PC terminals, and tablets;

[0030] AI, Artificial Intelligence;

[0031] The point areas include the school periphery, teaching areas, key areas and other areas; among them, the school periphery includes the school gate and perimeter walls; the teaching areas include classrooms, stairwells, corridors, stair entrances and exits, etc.; the key areas include rooftops, canteens, unmanned areas, sparsely populated areas, elevator cars and teachers' blind spots; the other areas include dormitories, libraries and computer rooms, etc.

[0032] The recognition sub-model includes various face recognition, body recognition, object recognition, behavior recognition, vehicle recognition, clothing recognition and other technologies, which are trained using algorithms such as artificial intelligence and deep learning (neural network).

[0033] In this embodiment, abnormal behavior, abnormal trajectory, abnormal number of people and abnormal list are used for explanation;

[0034] Facial cameras are deployed at the main road nodes on campus to provide early warning for suspicious persons on campus, thereby promptly detecting threat risks on campus.

[0035] This embodiment is described by taking the behavior analysis in analysis.

[0036] The behavioral anomalies are detected through a deep learning algorithm based on a neural network, which derives the human skeleton of each student in the video. According to the motion trajectory of the students’ limb movements, the coordinate changes of the skeleton in the picture are simulated, and the students’ behaviors are calculated based on the spatiotemporal characteristics of the motion change rules.

[0037] In different scenarios, different recognition sub-models of different scenario types are deployed to judge abnormal actions and behaviors. When students behave abnormally, an alarm is immediately pushed. During implementation, each scenario type can correspond to multiple recognition sub-models. One recognition sub-model corresponds to the associated recognition or analysis detection algorithm.

[0038] It should be noted that once the warning is triggered, the target person who triggered the warning will be captured immediately.

[0039] In this embodiment, by presetting more than 20 abnormal behavior analysis and detection algorithms such as fight detection, person presence detection, person crossing the boundary detection, and height climbing detection, the corresponding warnings generated include:

[0040] Fighting and bullying warning, wall climbing, truancy warning, rooftop warning, after-school delay warning, climbing over the door warning, falling down warning, gesture for help warning, illegal intrusion warning, single person warning, night stalking warning, takeaway detection warning, classroom discipline warning, burglary warning, dangerous climbing warning and abnormal gathering warning, etc.

[0041] For example, fight detection can be deployed in public areas, canteens, unmanned areas, sparsely populated areas, and teacher blind spots. Once campus bullying is detected, a fight and bullying warning will be issued immediately. At the same time, security guards, teachers, and school administrators will receive the warning and deal with it immediately.

[0042] The school perimeter wall is equipped with multiple detection algorithms to detect the presence of people, the retention of people, and the climbing of people on the wall. When a person is detected, an early warning is issued to remind the security guard to keep paying attention. When the target continues to stay, the alarm is upgraded. When the target climbs the wall, the security guard will step forward to deal with it.

[0043] The abnormal number of people is judged by combining the headcount of the facial access control channel gate, the in and out statistics of the dormitory entrance and exit cameras, the roll call statistics of the cameras in the classroom, the early warning data and the facial recognition technology;

[0044] Specifically, the facial recognition access control gate at the school gate can count how many students go to school every day, whether they are all present, and which class they are not from; after school, whether all students leave the school normally, and whether there are any students who are not left behind;

[0045] Students who ask for leave must go through the school gate. The number of students can be counted through the facial recognition access control gate, which does not count as an abnormal number of people. Only those who illegally climb over the wall to skip school will not be recorded in the system.

[0046] As well as capturing photos at key points such as dormitories, if there are students staying in the dormitory instead of attending school as usual, the cameras at the entrance and exit of the dormitory can be used to count the students entering and leaving the dormitory;

[0047] It is also possible to determine the abnormal number of students by associating it with abnormal warnings of truancy;

[0048] In the classroom, facial recognition technology is used to empower the cameras in the classroom to realize class roll call and determine if there are abnormalities in the number of students. Through the above-mentioned judgment of the number of people, the number of students can be counted in real time, which is equivalent to realizing seamless roll call.

[0049] Furthermore, it also includes a trajectory management module, which is used to generate a corresponding trajectory file for each student based on the structured data generated by each point area, combined with the time sequence, location and video surveillance data; that is, the corresponding trajectory is generated by extracting data such as personnel, time, location, photos, videos, etc. from various structured data;

[0050] When applied, a trajectory file is generated for each student every day; the corresponding trajectory status is divided into normal and abnormal, and the disposal status is pending and processed.

[0051] Furthermore, in order to achieve application flexibility, the trajectory management module is also used for trajectory control, which includes individual control, group control, regional control and time period control;

[0052] The personal control is used to accurately control a certain student;

[0053] The group control is used to control all objects within a certain range;

[0054] The area control is used to control the prohibited areas or the passable areas;

[0055] The time period control is used to control the time periods during which traffic is allowed or prohibited;

[0056] Among them, if there is control, once the student appears in the controlled area, the status is abnormal; if there is no control, the default status is normal.

[0057] The trajectory management module is also used for trajectory analysis, specifically:

[0058] The system automatically collects students' daily trajectory status. When a student's trajectory is abnormal for many consecutive times, corresponding warning data is generated and pushed to the guard on duty and the head teacher for warning reminders, so as to facilitate timely and proactive handling and reduce potential risks.

[0059] At the same time, the obtained early warning data is also used to optimize the aforementioned trajectory control to make it more targeted.

[0060] The trajectory management module is also used for target positioning. When a student's current location in school needs to be located in real time, the target positioning is used to retrieve the most recent snapshot address by searching with an image or name to locate the student's current location.

[0061] Furthermore, in order to further realize the combined application of personnel, equipment and maps, on the basis of the above-mentioned technical solution, the AI-based behavior analysis and recognition system also includes an association module, which is used to associate user information, personnel information, equipment information, terminal information and map information.

[0062] During implementation, user information corresponds to user management rules. Each user is associated with an organization. The user has the relevant permissions at the level of the organization.

[0063] There are personnel management rules corresponding to personnel information, and students and teachers can associate files according to organizational structures;

[0064] The device information corresponds to the device management rules. The locations of different devices are associated with organizational structures. After association, grouped preview of videos can be realized.

[0065] Terminal information corresponds to terminal management rules, mobile terminals are associated with PC terminals, and both are associated with organizational permissions;

[0066] Map information corresponds to map management rules. The campus master plan and floor layout map can be associated with organizational structures to achieve map navigation and map hierarchical management.

[0067] Furthermore, in this embodiment, the warning module is also used for warning management, and the warning management includes linkage warning, warning pop-up screen, message handling, alarm reporting, alarm timeout and multiple alarm warnings.

[0068] The linkage warning is to link the surrounding videos to display synchronously after the key area is alarmed;

[0069] The warning pop-up screen is pushed and the warning is issued immediately after the behavior analysis alarm;

[0070] The message handling is to handle the alarm by viewing the alarm information;

[0071] The alarm situation reporting means that when the alarm situation is serious, it can be manually reported to the command center;

[0072] The alarm timeout means that the command center will receive an alarm in time when the alarm is not handled in time;

[0073] The multiple alarm warnings are displayed in order of priority when multiple warnings are triggered at the same time.

[0074] Furthermore, the warning management also includes alarm strategy, alarm level definition, alarm escalation definition and alarm push definition.

[0075] During implementation, the rules for defining the alarm level are: the same event, occurring at different times and locations, may have very different severity and scope of impact. Take the appearance of people as an example: the toilet needs to be monitored for the presence of people. The purpose of the monitoring is to monitor every person who comes in to use the toilet. The danger is extremely low and only requires attention. However, if someone appears near the perimeter wall, next to the wall, or inside the wall, this is undoubtedly a very dangerous alarm.

[0076] The rules for defining police escalation are: when a police situation first occurs, the severity of the situation is different if you intervene immediately or if you allow it to develop without intervention;

[0077] The rules defined for alert push notifications are as follows: users can choose different push notification methods based on different states and situations such as alert levels and alert upgrades.

[0078] From the above description, it can be known that the technical solution provided by the embodiment of the present invention obtains video surveillance data collected by video surveillance equipment; sends the obtained video surveillance data to a pre-deployed AI analysis model to perform intelligent big data analysis to obtain corresponding analysis results; when the analysis results contain at least one preset abnormal behavior, triggers the generation of corresponding early warning data and pushes it; by empowering the monitoring system with AI technology, the system can timely discover relevant abnormal behaviors on campus through big data analysis, and actively warn and push; thereby improving the level of intelligence of campus safety management, changing campus safety from passive to active, reducing personnel negligence, and through timely and efficient processing, also reducing possible potential risks, thereby achieving the purpose of all-round security monitoring.

[0079] In the embodiments provided in the present application, it should be understood that the disclosed system can also be implemented in other ways. The system embodiments described above are merely schematic. It should also be noted that in some alternative implementations, the functions marked in the boxes can also occur in an order different from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0080] In addition, the functional modules in the various embodiments of the present invention can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part. When in use, each module only collects and stores user information with the full authorization of the user and in compliance with relevant laws and regulations, and protects the security and privacy of user data. Unauthorized access is strictly prohibited.

[0081] The processing of data will be carried out within the scope prescribed by law and shall not exceed the purpose and scope authorized by the user; at the same time, users shall enjoy relevant rights such as access, correction, deletion, restriction of processing, and rejection of their personal data; and strictly abide by applicable laws and regulations and conduct compliance reviews.

[0082] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0083] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or substitutions within the technical scope disclosed by the present invention, and these modifications or substitutions should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A behavior analysis and recognition system based on AI, characterized in that: The system comprises: The data input module is used to obtain the video surveillance data of each point area of ​​the smart campus collected by the video surveillance equipment; A behavior analysis module is used to send the obtained video surveillance data to a pre-deployed AI analysis model to perform intelligent big data analysis to obtain corresponding analysis results; wherein the AI ​​analysis model includes multiple recognition sub-models of different scene types; An early warning module is used to trigger the generation of corresponding early warning data when the analysis result contains at least one preset abnormal behavior; wherein the abnormal behavior includes abnormal behavior, abnormal trajectory, abnormal number of people and abnormal list; The alarm module is used to push the generated warning data.

2. The AI-based behavior analysis and recognition system according to claim 1, characterized in that: When the warning is triggered, the target person who triggered the warning is immediately captured.

3. The AI-based behavior analysis and recognition system according to claim 2, characterized in that: It also includes a trajectory management module, which is used to generate a corresponding trajectory file for each student based on the structured data generated by each point area, combined with the time sequence, location and video surveillance data.

4. The AI-based behavior analysis and recognition system according to claim 3, characterized in that: The trajectory management module is also used for trajectory control, which includes individual control, group control, regional control and time period control; Among them, if there is control, once the student appears in the controlled area, the status is abnormal; if there is no control, the default status is normal.

5. The AI-based behavior analysis and recognition system according to claim 4, characterized in that: The trajectory management module is also used for target positioning. When a student's current location in school needs to be located in real time, the target positioning is used to retrieve the most recent snapshot address by searching with an image or name to locate the student's current location.

6. The AI-based behavior analysis and recognition system according to claim 5, characterized in that: It also includes an association module, which is used to associate user information, associated personnel information, associated device information, associated terminal information and associated map information.

7. The AI-based behavior analysis and recognition system according to claim 6, characterized in that: The warning module is also used for warning management, which includes linkage warning, warning pop-up screen, message handling, alarm reporting, alarm timeout and multiple alarm warnings.

8. The AI-based behavior analysis and recognition system according to claim 7, characterized in that: The behavioral anomalies are detected through a deep learning algorithm based on a neural network, which derives the human skeleton of each student in the video. According to the motion trajectory of the students’ limb movements, the coordinate changes of the skeleton in the picture are simulated, and the students’ behaviors are calculated based on the spatiotemporal characteristics of the motion change rules.

9. The AI-based behavior analysis and recognition system according to claim 8, characterized in that: The abnormal number of people is judged by combining the number count of the facial access control channel gate, the entry and exit statistics of the dormitory entrance and exit cameras, the roll call statistics of the cameras in the classroom, early warning data and facial recognition technology.

Citation Information

Cited By

  • Intelligent pet drying box

    CN121420908A

  • Intelligent pet drying box

    CN224891519U