CN113158847A - Unsafe behavior detection method based on artificial intelligence - Google Patents
Unsafe behavior detection method based on artificial intelligence Download PDFInfo
- Publication number
- CN113158847A CN113158847A CN202110371164.5A CN202110371164A CN113158847A CN 113158847 A CN113158847 A CN 113158847A CN 202110371164 A CN202110371164 A CN 202110371164A CN 113158847 A CN113158847 A CN 113158847A
- Authority
- CN
- China
- Prior art keywords
- behavior
- safety
- unsafe
- human face
- behaviors
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Human Computer Interaction (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Computational Linguistics (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Alarm Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to an unsafe behavior detection method based on artificial intelligence, which comprises the following steps: step A1: acquiring and comparing the human face, carrying out human face snapshot on the agent in the monitoring area to form a head portrait, a whole body portrait and a short video with fixed time duration, carrying out human face snapshot library comparison on the human face snapshot in the monitoring area, and finding out whether the illegal action exists through the compared human face by virtue of the behavior feature library; a2, intelligently analyzing videos, judging whether uncertain behaviors such as boundary crossing, danger area entering and the like exist in real time, and recording and tracking the unknown behaviors; step A3: monitoring a contact behavior, defining behavior properties by combining the human face characteristics of the step A1 and the illegal behavior of the step A2, and judging whether the illegal behavior exists or not; step A4: environment safety detection, namely judging the safety condition of the environment through environment data; step A5: the unsafe behavior is further confirmed, suppressed and corrected by combining the violation behavior of step A3 and the environmental safety condition of step A4.
Description
Technical Field
The invention relates to the technical field of public safety, in particular to an unsafe behavior detection method based on artificial intelligence.
Background
The unsafe behavior refers to human errors causing human casualty accidents, including unsafe actions causing accidents and behaviors that should be done according to safety regulations but not done; unsafe behavior reflects the cause of the aspect of the person where the accident occurred.
The unsafe behaviors of people refer to human factors which influence safety and are expressed by people, and are behaviors which cause unnecessary damages to personnel and equipment and accidents when the people violate disciplines, operation procedures, methods and the like in the social activity process. The main causes of unsafe behavior and unsafe conditions in humans are: technical reasons, educational reasons, physical and attitude reasons, administrative reasons, etc.
The unsafe behaviors of people can be divided into conscious unsafe behaviors and unconscious unsafe behaviors, wherein the conscious unsafe behaviors refer to unsafe behaviors of purposeful, attempted and obvious deceased persons and are purposeful violation behaviors, and the unconscious unsafe behaviors refer to unsafe behaviors of unconsciousness and absence of needs and purposes.
The invention provides an artificial intelligence-based unsafe behavior detection method, which can effectively and automatically detect unsafe behaviors of people and reduce the workload of manual monitoring, thereby reducing safety accidents and reducing personal injuries.
Disclosure of Invention
The invention aims to provide an unsafe behavior detection method based on artificial intelligence, so as to solve the problems in the background technology.
In order to achieve the purpose, the invention provides the following technical scheme: an unsafe behavior detection method based on artificial intelligence comprises the following steps:
step A1: acquiring and comparing the human face, carrying out human face snapshot on the agent in the monitoring area to form a head portrait, a whole body portrait and a short video with fixed time duration, carrying out human face snapshot library comparison on the human face snapshot in the monitoring area, and finding out whether the illegal action exists through the compared human face by virtue of the behavior feature library;
a2, intelligently analyzing videos, judging whether uncertain behaviors such as boundary crossing, danger area entering and the like exist in real time, and recording and tracking the unknown behaviors;
step A3: monitoring a contact behavior, defining behavior properties by combining the human face characteristics of the step A1 and the illegal behavior of the step A2, and judging whether the illegal behavior exists or not;
step A4: environment safety detection, namely judging the safety condition of the environment through environment data;
step A5: the unsafe behavior is further confirmed, suppressed and corrected by combining the violation behavior of step A3 and the environmental safety condition of step A4.
Preferably, the human image acquisition and comparison comprises the steps of acquiring elements influencing campus safety by utilizing artificial intelligence, establishing a human face snapshot library, establishing a business form library, establishing an environmental state library, establishing a behavior characteristic library and carrying out safety management on the personal safety of teachers and students in the campus.
Preferably, the monitoring of the contact behavior comprises taking a suppression measure for conscious unsafe behavior, starting campus safety management, performing safety accident prevention, preventing accidents, and avoiding possible personal injuries to teachers and students; the safety education management is mainly carried out on unconscious unsafe behaviors, people are dissuaded from the safety education management, and meanwhile, the case push is used for guiding and enhancing the self-protection consciousness of students.
Preferably, the environmental safety detection comprises the following operation steps:
step B1: whether a fire disaster exists or not is checked, firstly, a smoke sensor is used for detecting a fixed point fire disaster, the fixed point fire disaster comprises a classroom, a library, a dining hall and the like, firstly, a camera is used for detecting smoke and fire, video analysis is carried out on a video captured by the camera, and the captured object mainly comprises a playground and a stadium, and once the smoke or the fire is found, dangerous behaviors of the fire disaster are predicted;
step B2: detecting whether heavy rain and heavy rain exist or not, and determining whether heavy rain exists or not through a rainfall detector;
step B3: detecting whether the temperature is high or not, detecting the ambient temperature once every one minute by a temperature sensor, including indoor and outdoor, and if the temperature exceeds a specified temperature, determining that a high-temperature ambient state exists;
step B4: the traffic congestion or the personnel gathering condition is detected, the behavior analysis is carried out through the video uploaded by the camera, and for key points such as school gates, if the number of behavior people counted in the delineating area exceeds the specified number of people, the situation that the personnel gathering behaviors exist is indicated.
Preferably, a detection process is completed, then the system carries out environmental safety evaluation, if one environmental safety or multiple safety behaviors exist, the environment is judged to have one or multiple safety problems, and the early warning level of the whole safety condition is increased.
The invention has the technical effects and advantages that:
1. the invention has the functions of automatically detecting the agent, identifying the unsafe behavior, early warning the unsafe behavior and providing a basis for controlling the unsafe behavior.
2. The invention is beneficial to discovering potential safety hazards and sources, avoids safety injury accidents and ensures the safety of public places such as campuses, parks and the like.
3. The invention can effectively reduce the probability of safety accidents in dense areas such as campuses and the like, early warns, effectively ensures the safety of people and is convenient for wide popularization and use.
Drawings
Fig. 1 is a flowchart of an unsafe behavior detection method based on artificial intelligence according to the present invention.
Fig. 2 is a sequence diagram of an unsafe behavior detection method based on artificial intelligence according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention provides an unsafe behavior detection method based on artificial intelligence, which is shown in figures 1-2 and comprises the following steps:
step A1: and acquiring and comparing the human faces, carrying out face snapshot on the actors in the monitoring area to form a head portrait, a whole body portrait and a short video with fixed time duration, carrying out face snapshot library comparison on the faces snapshot in the monitoring area, finding out whether the violations exist through the compared faces through a behavior feature library, and sequentially arranging the violations according to the similarity.
The face that will grab is preserved in the face storehouse simultaneously, no matter this time the face of grabbing has or not, prepare for subsequent face identification and comparison, grab and shoot face information and make the record, because the face represents the characteristic of people, the face of video snapshot is deposited in the face storehouse, the face quantity that the face storehouse was stored is more, face identification's the degree of accuracy is higher, for same person, the face quantity that shoots through different angles, different places, under the different illumination condition is more, its facial characteristic is abundanter, therefore, when discerning this action person, the detection rate and the degree of accuracy of discernment are very close to this action person.
The human faces in the human face library not only have human face characteristics, but also store personnel information related to the human faces. In addition, a face information system is accessed through a network, and the identity information of the agent is inquired through the face.
The human face is obtained from continuous video streaming, namely, the human face is dynamically recognized, the personnel in the monitored area are dynamically recorded, the personnel captured in the monitored area are dynamically counted, once an emergency such as a fire occurs, the accurate number of the personnel in the area can be immediately and accurately given, the condition of the personnel in the monitored area is found through the human face characteristics, and a basis is provided for emergency rescue.
And step A2, intelligently analyzing videos, judging whether the uncertain behaviors such as boundary crossing, danger area entering and the like are violated in real time, and recording and tracking the unknown behaviors.
The campus equipment has potential safety hazards, most middle and primary schools put main efforts on teachers and students in the working process of safety management, but neglect campus environment and campus facility problems and some external threats, and behavioral analysis is to frame teaching facilities and places with certain potential safety hazards in areas, students entering the places are possibly injured by safety and even influence the life safety of the students, and the method is used for monitoring and preventing the students from entering or approaching dangerous facilities and places and reducing safety incidents.
The monitoring measures of the behavior analysis are that manual line drawing is carried out on a visual field area of a camera through framing of a monitored area to form a safety line, and behaviors crossing the safety line are regarded as dangerous behaviors.
The behavior analysis is used for monitoring the areas inside the campus, mainly detecting dangerous facilities which are damaged and in a state to be repaired, if the facilities are continuously used, personal injuries of students can be caused, the areas where the facilities are located prohibit the students from approaching, and the intelligent video analysis is mainly used for judging whether a person breaks into the areas or not and finding out border-crossing behaviors.
Behavioral analysis is to the monitoring of campus outside area, and the gate of campus is the key area of concern, and this will effectively protect student's personal safety, prevents that the injury incident from taking place.
And (4) behavior analysis processes the video stream in real time, finds out the illegal behaviors from the video, dynamically records the illegal behaviors, dynamically counts the illegal behaviors, and finds out the face characteristic information of the illegal actor.
Step A3: and (4) monitoring the contact behavior, combining the human face characteristics of the step A1 and the violation behavior of the step A2, defining behavior properties, and judging whether the violation behavior exists.
At present, many middle and primary schools are built in city centers or busy places in towns, so that the environment around the campus is relatively complex, although school managers strengthen supervision and management on students in the school, school managers can hardly control and manage students outside the school, and therefore, the detection of unsafe behaviors in advance is very important for avoiding safety incidents.
For pupils, the purpose of getting on and off school is a key link, and strangers pretend to be parents of illegal behaviors should be warned in advance.
Monitoring a behavior event contacting with a student, if the behavior event occurs in an internal area of a campus, defining the behavior as a benign behavior, if the behavior event occurs in an external area of the campus, determining the behavior as an inferior behavior, simultaneously calling a captured face of an agent, and if the face of the agent does not exist in a face library, immediately generating an alarm notification; if the human face is stored in the action library, the action record of the action in the action library is called through the human face, the previous action and the current action of the action are integrated for evaluation, if the evaluation is poor, an alarm notification immediately occurs, and after the receiving and sending responsible person receives the alarm notification, the action is confirmed, if the action is a parent, the alarm notification is cancelled, and if not, security personnel are notified to take isolation measures.
Step A4: and (4) environment safety detection, namely judging the safety condition of the environment through environment data.
In a campus of middle and small schools, accidents threaten students, teachers and students cannot predict the accidents, the accidents not only cause huge economic loss to schools, but also threaten lives of the students and the teachers, such as fire, heavy rain, high temperature and other natural factors, which are important factors influencing campus safety management, and therefore monitoring and evaluating the overall safety factors of the campuses, namely the safety conditions of the environments where the students are located, are necessary.
The environmental safety detection is implemented according to the following steps:
step B1: whether the inspection conflagration exists, firstly detect the fixed point conflagration through smoke transducer, the fixed point includes classroom, library, dining room etc. secondly detects smog and flame through the camera, carries out video analysis to the video that the camera was grabbed and is clapped, and the object of clapping mainly is playground, stadium, no matter which kind of mode, in case when discovering smog or flame, indicates promptly that there is the dangerous behavior of conflagration.
Step B2: and detecting whether heavy rain and heavy rain exist or not, and determining whether heavy rain exists or not through a rainfall detector.
Step B3: whether the temperature is high or not is detected, the ambient temperature is detected once every minute by a temperature sensor, including indoor and outdoor, and if the temperature exceeds a specified temperature, the high-temperature ambient state is considered to exist.
Step B4: the traffic congestion or the personnel gathering condition is detected, the behavior analysis is carried out through the video uploaded by the camera, and for key points such as school doors, if the number of behavior people counted in the delineating area exceeds the specified number of people, the situation that the personnel gathering behavior exists is indicated.
Once a detection process is completed, the system carries out environmental safety evaluation immediately, if one environmental safety or multiple safety behaviors exist, the environment is judged to have one or multiple safety problems, and the early warning level of the whole safety condition is increased.
Step A5: the unsafe behavior is further confirmed, suppressed and corrected by combining the violation behavior of step A3 and the environmental safety condition of step A4.
From the behavior versus environment analysis, the same behavior, in one environment, is a safe behavior, in another environment, is an unsafe behavior, and the confirmation of the unsafe behavior, considering the safety condition of the environment, is an unsafe behavior if the traffic congestion is detected in step 4 and the violation occurs in step 3.
The unconscious unsafe behavior refers to the unsafe behavior without need and purpose, which is random in statistical analysis, and the unsafe behavior of people is the direct reason of accidents.
The method determines the dangerous behaviors of an agent by using a statistical method, gives the random degree of the behaviors of the agent, further judges whether the dangerous behaviors are unconscious unsafe behaviors or conscious unsafe behaviors, and inhibits or corrects the dangerous behaviors in a targeted manner.
Restraining measures are taken for conscious unsafe behaviors, campus safety management is started, safety accident prevention is carried out, accidents are prevented, and possible personal injuries to teachers and students are avoided; the safety education management is mainly carried out on unconscious unsafe behaviors, people are dissuaded from the safety education management, and meanwhile, the case push is used for guiding and enhancing the self-protection consciousness of students.
In the embodiment, the elements affecting the campus safety are collected by using artificial intelligence, a face snapshot library is established, a list library is established, an environment state library is established, a behavior characteristic library is established, and the personal safety of teachers and students in the campus is managed.
In the embodiment, the step of monitoring the contact behavior comprises the steps of taking a restraining measure for conscious unsafe behaviors, starting campus safety management, preventing safety accidents, preventing accidents and avoiding possible personal injuries to teachers and students; the safety education management is mainly carried out on unconscious unsafe behaviors, people are dissuaded from the safety education management, and meanwhile, the case push is used for guiding and enhancing the self-protection consciousness of students.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments or portions thereof without departing from the spirit and scope of the invention.
Claims (5)
1. An unsafe behavior detection method based on artificial intelligence is characterized by comprising the following steps:
step A1: acquiring and comparing the human face, carrying out human face snapshot on the agent in the monitoring area to form a head portrait, a whole body portrait and a short video with fixed time duration, carrying out human face snapshot library comparison on the human face snapshot in the monitoring area, and finding out whether the illegal action exists through the compared human face by virtue of the behavior feature library;
a2, intelligently analyzing videos, judging whether uncertain behaviors such as boundary crossing, danger area entering and the like exist in real time, and recording and tracking the unknown behaviors;
step A3: monitoring a contact behavior, defining behavior properties by combining the human face characteristics of the step A1 and the illegal behavior of the step A2, and judging whether the illegal behavior exists or not;
step A4: environment safety detection, namely judging the safety condition of the environment through environment data;
step A5: the unsafe behavior is further confirmed, suppressed and corrected by combining the violation behavior of step A3 and the environmental safety condition of step A4.
2. The artificial intelligence based unsafe behavior detection method of claim 1, wherein: the portrait acquisition comparison comprises the steps of acquiring elements influencing campus safety by utilizing artificial intelligence, establishing a face snapshot library, establishing a list library, establishing an environmental state library, establishing a behavior characteristic library, and carrying out safety management on the personal safety of teachers and students in the campus.
3. The artificial intelligence based unsafe behavior detection method of claim 1, wherein: the contact behavior monitoring comprises the steps of taking inhibition measures for conscious unsafe behaviors, starting campus safety management, preventing safety accidents, preventing accidents and avoiding possible personal injuries to teachers and students; the safety education management is mainly carried out on unconscious unsafe behaviors, people are dissuaded from the safety education management, and meanwhile, the case push is used for guiding and enhancing the self-protection consciousness of students.
4. The artificial intelligence based unsafe behavior detection method of claim 1, wherein: the environmental security detection comprises the following operation steps:
step B1: whether a fire disaster exists or not is checked, firstly, a smoke sensor is used for detecting a fixed point fire disaster, the fixed point fire disaster comprises a classroom, a library, a dining hall and the like, firstly, a camera is used for detecting smoke and fire, video analysis is carried out on a video captured by the camera, and the captured object mainly comprises a playground and a stadium, and once the smoke or the fire is found, dangerous behaviors of the fire disaster are predicted;
step B2: detecting whether heavy rain and heavy rain exist or not, and determining whether heavy rain exists or not through a rainfall detector;
step B3: detecting whether the temperature is high or not, detecting the ambient temperature once every one minute by a temperature sensor, including indoor and outdoor, and if the temperature exceeds a specified temperature, determining that a high-temperature ambient state exists;
step B4: the traffic congestion or the personnel gathering condition is detected, the behavior analysis is carried out through the video uploaded by the camera, and for key points such as school gates, if the number of behavior people counted in the delineating area exceeds the specified number of people, the situation that the personnel gathering behaviors exist is indicated.
5. The artificial intelligence based unsafe behavior detection method of claim 4, wherein: and completing a detection process, then carrying out environmental safety evaluation by the system, and if one environmental safety or multiple safety behaviors exist, judging that one or multiple safety problems exist in the environment, and increasing the early warning level of the whole safety condition.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110371164.5A CN113158847A (en) | 2021-04-07 | 2021-04-07 | Unsafe behavior detection method based on artificial intelligence |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110371164.5A CN113158847A (en) | 2021-04-07 | 2021-04-07 | Unsafe behavior detection method based on artificial intelligence |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113158847A true CN113158847A (en) | 2021-07-23 |
Family
ID=76888508
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110371164.5A Pending CN113158847A (en) | 2021-04-07 | 2021-04-07 | Unsafe behavior detection method based on artificial intelligence |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113158847A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116127786A (en) * | 2023-04-07 | 2023-05-16 | 南京大学 | System and method for measuring and calculating security state of slow traffic group based on city slow traffic simulation |
-
2021
- 2021-04-07 CN CN202110371164.5A patent/CN113158847A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116127786A (en) * | 2023-04-07 | 2023-05-16 | 南京大学 | System and method for measuring and calculating security state of slow traffic group based on city slow traffic simulation |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7595815B2 (en) | Apparatus, methods, and systems for intelligent security and safety | |
CN109413369A (en) | A kind of monitor video intellectual analysis early warning platform | |
CN106454250A (en) | Intelligent recognition and early warning processing information platform | |
Ogunleye et al. | A computer-based security framework for crime prevention in Nigeria | |
WO2021023198A1 (en) | Train safety driving monitoring system and method | |
WO2021023064A1 (en) | Safe driving monitoring system and method for train | |
CN110245630A (en) | Monitoring data processing method, device and readable storage medium storing program for executing | |
CN101674461A (en) | Intelligent network monitoring system for safety of primary and secondary school campuses | |
CN108765921A (en) | View-based access control model lexical analysis is applied to the intelligent patrol method of patrol robot | |
CN112349015B (en) | Shelter management and control system and method based on Internet of things | |
CN111985428A (en) | Security detection method and device, electronic equipment and storage medium | |
CN113158847A (en) | Unsafe behavior detection method based on artificial intelligence | |
CN111681355A (en) | Security and protection system applicable to smart campus | |
CN107920224B (en) | Abnormity warning method, abnormity warning equipment and video monitoring system | |
CN112211496A (en) | Monitoring method and system based on intelligent door lock and intelligent door lock | |
CN111553264B (en) | Campus non-safety behavior detection and early warning method suitable for primary and secondary school students | |
CN111820880A (en) | Campus overlord early warning system and method | |
KR102648004B1 (en) | Apparatus and Method for Detecting Violence, Smart Violence Monitoring System having the same | |
WO2023158926A1 (en) | Systems and methods for detecting security events in an environment | |
CN115762046A (en) | Early warning and intervention device for building top light life person | |
CN115376268A (en) | Monitoring alarm fire-fighting linkage system based on image recognition | |
CN113705357A (en) | Method, system, device and storage medium for identifying electric vehicle based on camera | |
CN209199284U (en) | A kind of Household security system | |
CN114694344A (en) | Campus violence monitoring method and device and electronic equipment | |
CN106408855A (en) | Intelligent environment management and control system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |