CN115294515A - Artificial intelligence-based comprehensive anti-theft management method and system - Google Patents

Artificial intelligence-based comprehensive anti-theft management method and system Download PDF

Info

Publication number
CN115294515A
CN115294515A CN202210793697.7A CN202210793697A CN115294515A CN 115294515 A CN115294515 A CN 115294515A CN 202210793697 A CN202210793697 A CN 202210793697A CN 115294515 A CN115294515 A CN 115294515A
Authority
CN
China
Prior art keywords
image
determining
human body
cell
person
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.)
Granted
Application number
CN202210793697.7A
Other languages
Chinese (zh)
Other versions
CN115294515B (en
Inventor
章韵
褚富强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Posts and Telecommunications
Original Assignee
Nanjing University of Posts and Telecommunications
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Nanjing University of Posts and Telecommunications filed Critical Nanjing University of Posts and Telecommunications
Priority to CN202210793697.7A priority Critical patent/CN115294515B/en
Publication of CN115294515A publication Critical patent/CN115294515A/en
Application granted granted Critical
Publication of CN115294515B publication Critical patent/CN115294515B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/62—Extraction of image or video features relating to a temporal dimension, e.g. time-based feature extraction; Pattern tracking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • G06V10/752—Contour matching
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • 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
    • G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion
    • 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
    • G06V40/103—Static body considered as a whole, e.g. static pedestrian or occupant recognition
    • 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
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
    • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
    • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
    • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
    • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
    • G08B13/19608—Tracking movement of a target, e.g. by detecting an object predefined as a target, using target direction and or velocity to predict its new position

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Human Computer Interaction (AREA)
  • Artificial Intelligence (AREA)
  • Databases & Information Systems (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Social Psychology (AREA)
  • Psychiatry (AREA)
  • Image Analysis (AREA)
  • Alarm Systems (AREA)

Abstract

The invention provides a comprehensive anti-theft management method and a comprehensive anti-theft management system based on artificial intelligence, wherein the method comprises the following steps: acquiring a first image of a person entering a cell; determining suspicious people based on the first image and a preset portrait library; tracking and acquiring a second image of the suspicious person in the cell; and determining the theft behavior and outputting an alarm based on the second image and a preset neural network model. The comprehensive anti-theft management method based on artificial intelligence intelligently tracks suspicious personnel entering the community intelligently, analyzes the behavior of the personnel, further realizes intelligent discovery of the suspicious personnel, stops the theft behavior in the community, and practically ensures the property safety of owners in the community.

Description

Artificial intelligence-based comprehensive anti-theft management method and system
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to an artificial intelligence-based comprehensive anti-theft management method and system.
Background
At present, with the advance of urbanization, most people live in a cell; the property of district is responsible for cleaning, the safe work in district, and the theftproof of district is mainly through laying the surveillance camera head in the district essential position, looks over the mode of the picture that the surveillance camera head was shot through the manual work, and efficiency is lower to can not in time discover the theft action, often be the theft after taking place, select suspicious personnel through the picture of surveillance camera head.
Disclosure of Invention
One of the purposes of the invention is to provide a comprehensive anti-theft management method based on artificial intelligence, which can intelligently track suspicious people entering a community and analyze the behaviors of the people, so as to intelligently discover the suspicious people and terminate the theft behavior in the community, thereby practically ensuring the property safety of owners in the community.
The embodiment of the invention provides a comprehensive anti-theft management method based on artificial intelligence, which comprises the following steps:
acquiring a first image of a person entering a cell;
determining suspicious people based on the first image and a preset portrait library;
tracking and acquiring a second image of the suspicious person in the cell;
and determining the theft behavior and outputting an alarm based on the second image and a preset neural network model.
Preferably, the acquiring of the first image of the person entering the cell comprises:
acquiring a first image of a person entering a cell by a law enforcement instrument worn by a security guard at a cell doorway;
and/or the presence of a gas in the gas,
a first image of a person entering a cell is acquired through a first image acquisition device arranged beside an entrance and an exit of the cell.
Preferably, the method for determining suspicious people based on the first image and a preset portrait base comprises the following steps:
extracting the human body contour of the first image;
when the number of the extracted human body outlines is one, matching the images corresponding to the human body outlines with all the human images in the human image library, and determining whether the people corresponding to the human body outlines are suspicious people;
when the number of the extracted human body contours is more than one, determining the relative positions of the human body contours and whether an overlapping area exists or not;
extracting features of the relative position and the overlapping area based on a preset first feature extraction template to obtain a plurality of feature values;
constructing a relation description vector between human body contours based on the characteristic values;
acquiring a preset relation judgment library;
determining the relationship between the human body contours based on the relationship judgment library and the relationship description vector;
grouping the extracted human body contours based on the relationship between the human body contours to obtain a plurality of groups;
extracting any human body contour of each group according to a preset extraction rule, matching an image corresponding to the human body contour with each human image in a human image library, and determining whether a person corresponding to the human body contour is a suspicious person or not, wherein when the person corresponding to any human body contour in the same group is not the suspicious person, other persons in the same group are not the suspicious person;
wherein, the portrait base is constructed by the following steps:
acquiring a third image of each owner in the cell;
extracting a portrait of the third image;
and/or the presence of a gas in the atmosphere,
acquiring historical records of cell access;
analyzing the historical records, and determining fourth images of all people of which the times of entering and exiting the cell in a preset time period reach a preset time threshold;
and extracting the portrait of the fourth image.
Preferably, the tracking and acquiring the second image of the suspicious person in the cell includes:
constructing a virtual map of a monitoring facility of a cell;
acquiring the position, the moving direction and the moving speed of the suspicious personnel;
based on the position of the suspicious person, mapping the suspicious person to a virtual map;
determining a second image acquisition device arranged in the cell for tracking shooting and a corresponding shooting start time for determining a second image based on the moving direction, the moving speed and the virtual map;
acquiring a second image of the suspicious person through second image acquisition equipment;
wherein determining a second image capturing device disposed within the cell for tracking shooting and a corresponding shooting start time for determining the second image based on the moving direction, the moving speed, and the virtual map comprises:
determining a first direction vector based on the moving direction and the position of the suspicious person in the virtual map;
determining a plurality of second direction vectors based on the position of the suspicious person in the virtual map and the setting position of each second image acquisition device;
respectively calculating the included angle between each second direction vector and the first direction vector;
extracting second image acquisition equipment corresponding to a second direction vector with an included angle smaller than a preset threshold value;
determining the distance between each extracted second image acquisition device and the position of the suspicious person;
and the second image acquisition device with the minimum distance is used as the second image acquisition device for tracking and shooting.
Preferably, the determining the theft and outputting the alarm based on the second image and the preset neural network model comprises:
extracting the features of the second image based on a preset second feature extraction template to obtain a plurality of behavior feature values;
inputting a plurality of behavior characteristic values into a neural network model, and determining a behavior analysis result;
and outputting an alarm when the behavior analysis result is the theft behavior.
The invention also provides a comprehensive anti-theft management system based on artificial intelligence, which comprises:
the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring a first image of a person entering a cell;
the first determination module is used for determining suspicious people based on the first image and a preset portrait library;
the second acquisition module is used for tracking and acquiring a second image of the suspicious personnel in the cell;
and the second determination module is used for determining the theft behavior and outputting an alarm based on the second image and a preset neural network model.
Preferably, the first acquiring module acquires a first image of a person entering the cell, and performs the following operations:
acquiring a first image of a person entering a cell by a law enforcement instrument worn by a security guard at a cell doorway;
and/or the presence of a gas in the gas,
the method comprises the steps of obtaining a first image of a person entering a community through first image collecting equipment arranged beside an entrance and an exit of the community.
Preferably, the first determination module determines the suspicious person based on the first image and a preset portrait base, and performs the following operations:
extracting the human body contour of the first image;
when the number of the extracted human body contours is one, matching the images corresponding to the human body contours with the human images in the human image library to determine whether the people corresponding to the human body contours are suspicious people;
when the number of the extracted human body contours is more than one, determining the relative positions of the human body contours and whether an overlapping area exists or not;
performing feature extraction on the relative position and the overlapping area based on a preset first feature extraction template to obtain a plurality of feature values;
constructing a relation description vector between the human body contours based on the characteristic values;
acquiring a preset relation judgment library;
determining the relationship between the human body contours based on the relationship judgment library and the relationship description vector;
grouping the extracted human body contours based on the relationship between the human body contours to obtain a plurality of groups;
extracting any human body contour of each group according to a preset extraction rule, matching an image corresponding to the human body contour with each human image in a human image library, and determining whether a person corresponding to the human body contour is a suspicious person or not, wherein when the person corresponding to any human body contour in the same group is not the suspicious person, other persons in the same group are not the suspicious person;
wherein, the portrait bank is constructed by the following steps:
acquiring a third image of each owner in the cell;
extracting a portrait of the third image;
and/or the presence of a gas in the gas,
acquiring historical records of cell access;
analyzing the historical records, and determining fourth images of all people of which the times of entering and exiting the cell in a preset time period reach a preset time threshold;
and extracting the portrait of the fourth image.
Preferably, the second acquiring module tracks and acquires a second image of the suspicious person in the cell, and performs the following operations:
constructing a virtual map of a monitoring facility of a cell;
acquiring the position, the moving direction and the moving speed of the suspicious personnel;
mapping the suspicious person to a virtual map based on the location of the suspicious person;
determining a second image acquisition device arranged in the cell for tracking shooting and a corresponding shooting start time for determining a second image based on the moving direction, the moving speed and the virtual map;
acquiring a second image of the suspicious person through second image acquisition equipment;
wherein determining a second image capturing device disposed within the cell for tracking shooting and a corresponding shooting start time for determining the second image based on the moving direction, the moving speed, and the virtual map comprises:
determining a first direction vector based on the moving direction and the position of the suspicious person in the virtual map;
determining a plurality of second direction vectors based on the position of the suspicious person in the virtual map and the setting position of each second image acquisition device;
respectively calculating the included angle between each second direction vector and the first direction vector;
extracting second image acquisition equipment corresponding to a second direction vector with an included angle smaller than a preset threshold value;
determining the distance between each extracted second image acquisition device and the position of the suspicious person;
and the second image acquisition device with the minimum distance is used as the second image acquisition device for tracking and shooting.
Preferably, the second determining module determines the theft and outputs an alarm based on the second image and a preset neural network model, and performs the following operations:
extracting the features of the second image based on a preset second feature extraction template to obtain a plurality of behavior feature values;
inputting a plurality of behavior characteristic values into a neural network model, and determining a behavior analysis result;
and outputting an alarm when the behavior analysis result is the theft behavior.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
FIG. 1 is a schematic diagram of an integrated anti-theft management method based on artificial intelligence in an embodiment of the present invention;
fig. 2 is a schematic diagram of an artificial intelligence based integrated antitheft management system according to an embodiment of the present invention.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it should be understood that they are presented herein only to illustrate and explain the present invention and not to limit the present invention.
The embodiment of the invention provides a comprehensive anti-theft management method based on artificial intelligence, which comprises the following steps as shown in figure 1:
step S1: acquiring a first image of a person entering a cell;
step S2: determining suspicious people based on the first image and a preset portrait library;
and step S3: tracking and acquiring a second image of the suspicious person in the cell;
and step S4: and determining the theft behavior and outputting an alarm based on the second image and a preset neural network model.
The working principle and the beneficial effects of the technical scheme are as follows:
acquiring a first image through equipment when a person enters a cell, and determining whether the person is a person corresponding to a portrait in a portrait library through a portrait library so as to determine whether the person is a suspicious person; the method comprises the steps of chasing up and shooting determined suspicious people, outputting a shot second image to a preset neural network model, judging whether the behavior of the suspicious people is a theft behavior intelligently, and outputting an alarm to community workers at the doorway of a community, so that the theft behavior is stopped in the community, and the property safety of residents of the community is improved, wherein the neural network model is an artificial intelligent network model which is trained and converged based on images of the behaviors of the theft people in a large number of theft events.
The comprehensive anti-theft management method based on artificial intelligence intelligently tracks suspicious personnel entering the community intelligently, analyzes the behavior of the personnel, further realizes intelligent discovery of the suspicious personnel, stops the theft behavior in the community, and practically ensures the property safety of owners in the community.
In one embodiment, acquiring a first image of a person entering a cell comprises:
acquiring a first image of a person entering a cell by a law enforcement instrument worn by a security guard at a cell doorway;
and/or the presence of a gas in the atmosphere,
a first image of a person entering a cell is acquired through a first image acquisition device arranged beside an entrance and an exit of the cell.
The working principle and the beneficial effects of the technical scheme are as follows:
the law enforcement instrument is worn in front of the chest for the security officer working at the entrance of the community, and is used for shooting the law enforcement situation of the security officer in the community and shooting personnel entering the community from the entrance of the community; the first image pickup apparatus includes: shooting equipment such as a high-level camera arranged at the doorway of the community; the first image of each person entering the community is guaranteed to be shot through the comprehensive law enforcement instrument and the first image acquisition device.
In one embodiment, determining suspicious people based on the first image and a preset portrait library comprises:
extracting the human body contour of the first image;
when the number of the extracted human body outlines is one, matching the images corresponding to the human body outlines with all the human images in the human image library, and determining whether the people corresponding to the human body outlines are suspicious people; that is, only one person enters the cell at the same time; only suspicious personnel verification is needed to be carried out on the personnel;
when the number of the extracted human body contours is more than one, determining the relative positions of the human body contours and whether an overlapping area exists; namely, a plurality of persons enter a cell at the same time, and the relationship among the persons is judged firstly; for example: the old people enter a cell by pulling hands of children, so that the outlines of the two human bodies are overlapped, and lovers, couples and the like can be realized; in addition, the distance between the walking of the people with close relation is closer than that of the strangers; therefore, whether the entering personnel know each other can be effectively judged by judging whether the relative position is overlapped with the human body outline or not;
extracting features of the relative position and the overlapping area based on a preset first feature extraction template to obtain a plurality of feature values; the characteristic values include: a distance value indicating a relative position, a position of the overlap region in the human body contour, an area of the overlap region, and the like;
constructing a relation description vector between the human body contours based on the characteristic values; orderly arranging the characteristic values to obtain a relation description vector;
acquiring a preset relation judgment library; the relations between the relation description vectors and the human body contours in the relation judgment library are in one-to-one corresponding correlation;
determining the relationship between the human body contours based on the relationship judgment library and the relationship description vector; the constructed relation description vector is matched with the relation description vector in the relation judgment library, and when the matching is consistent, the corresponding correlation relation of the relation description vector in the relation judgment library is obtained;
grouping the extracted human body contours based on the relationship between the human body contours to obtain a plurality of groups; the human body outlines of acquainted people are divided into a group;
extracting any human body contour of each group according to a preset extraction rule (sequentially or randomly), matching the image corresponding to the human body contour with each human image in a human image library (for example, calculating the similarity between the image corresponding to the human body contour and each human image, and when the similarity is greater than a preset threshold value, determining that the image and the human image are matched), determining whether the personnel corresponding to the human body contour are suspicious personnel, and when the personnel corresponding to any human body contour in the same group are not suspicious personnel, determining that other personnel in the same group are not suspicious personnel; therefore, the human body outlines of the same group do not need to be matched with the human images in the human image library, and the judgment of the suspicious personnel of the group can be finished as long as one human body outline is found to be matched with the human images in the human image library in the process of extracting and matching, wherein the group is not the suspicious personnel; furthermore, theft generally does not exceed 3 people, so when the number of people in the same group is greater than 3 people, it can be excluded as a suspect;
wherein, the portrait bank is constructed by the following steps:
acquiring a third image of each owner in the cell; perfecting a portrait base by registering owners in a cell;
extracting a portrait of the third image; extracting the human body contour of the third image;
and/or the presence of a gas in the gas,
acquiring historical records of cell access; the historical record is a first image acquired by a cell entrance law enforcement instrument or first image acquisition equipment;
analyzing the historical records, and determining a fourth image of each person of which the number of times of entering and exiting the cell reaches a preset number threshold (for example, two times) in a preset time period (for example, within 1 day);
and extracting the portrait of the fourth image.
In one embodiment, tracking and acquiring a second image of the suspicious person in the cell comprises:
constructing a virtual map of monitoring facilities of a cell; the virtual map is a plan view of the cell, and the positions of the monitoring facilities (second image acquisition equipment) are marked on the plan view;
acquiring the position, the moving direction and the moving speed of the suspicious personnel; for the person walking normally, image recognition analysis can be carried out through two first images separated by N frames, and the moving direction and the moving speed of the suspicious person are determined; determining the moving speed by dividing the moving distance of the suspicious people of the previous frame and the next frame by the time difference between the two frames; the moving direction is the position from the previous frame to the next frame; determining the position of the suspicious person, taking the acquisition of a law enforcement instrument as an example, acquiring the positioning information of the law enforcement instrument through the law enforcement instrument, and determining the position of the suspicious person relative to the law enforcement instrument by carrying out image recognition on the first image so as to further determine the position of the suspicious person; taking a high-level camera as an example, knowing the installation position of the high-level camera, and determining the position of the suspicious person relative to the high-level camera by performing image recognition on the first image so as to determine the position of the suspicious person;
mapping the suspicious person to a virtual map based on the location of the suspicious person; positions on the virtual map correspond to real positions in the real cell one by one;
determining a second image acquisition device arranged in the cell for tracking shooting and a corresponding shooting start time for determining a second image based on the moving direction, the moving speed and the virtual map;
acquiring a second image of the suspicious person through second image acquisition equipment;
wherein determining a second image capturing device disposed in the cell for tracking shooting and a corresponding shooting start time for determining the second image based on the moving direction, the moving speed and the virtual map comprises:
determining a first direction vector based on the moving direction and the position of the suspicious person in the virtual map; the first direction vector is a unit vector of which the direction of the position of the suspicious person in the virtual map is consistent with the moving direction;
determining a plurality of second direction vectors based on the position of the suspicious person in the virtual map and the setting position of each second image acquisition device; the second direction vector points to the setting position of the second image acquisition equipment from the position of the suspicious person in the virtual map;
respectively calculating the included angle between each second direction vector and the first direction vector; the calculation of the included angle can adopt a vector included angle calculation formula, and the formula is as follows:
Figure BDA0003731340050000101
wherein x is 1 ，y 1 Respectively an abscissa and an ordinate of the first direction vector; x is a radical of a fluorine atom 2 ，y 2 Respectively an abscissa and an ordinate of the second direction vector; theta is an included angle;
extracting a second image acquisition device corresponding to a second direction vector with an included angle smaller than a preset threshold (for example, 5 degrees);
determining the distance between each extracted second image acquisition device and the position of the suspicious person;
taking the second image acquisition equipment with the minimum distance as second image acquisition equipment for tracking and shooting;
determining a photographing start time based on the value of the distance with the minimum distance and the moving speed; the shooting start time is a value of the distance divided by the moving speed.
In one embodiment, determining the theft and outputting an alarm based on the second image and a preset neural network model comprises:
extracting the features of the second image based on a preset second feature extraction template to obtain a plurality of behavior feature values; the behavior feature values include: whether the hand holds an article or not, whether the hand carries a backpack or not, whether the hand is walking with a head down or not, walking speed, whether the hand shoots a face or not, whether the hand enters a residential building or not, time difference between the hand and the residential building after entering the residential building again and the like;
inputting a plurality of behavior characteristic values into a neural network model, and determining a behavior analysis result;
and outputting an alarm when the behavior analysis result is the theft behavior.
The behavior of suspicious people in the second image is analyzed through the neural network model, whether the second image is a theft or not is determined, intelligent analysis of the theft behavior is achieved, and the intelligent analysis is more stable compared with manual monitoring.
In one embodiment, the comprehensive anti-theft management method based on artificial intelligence further comprises:
receiving an access application of a third image acquisition device installed by a community owner;
acquiring a setting position of third image acquisition equipment;
adding a third image capture device into the virtual map.
The working principle and the beneficial effects of the technical scheme are as follows:
through installing the district owner in the third image acquisition equipment access district's at the door and entrance, will track suspicious personnel and extend to the district owner, will prevent burglary and extend to each owner's family, further guaranteed owner's property safety.
In order to further ensure the property safety of the owners, in one embodiment, a white list is established for each third image acquisition device, and the personnel in the white list are input for the corresponding owners; and when the third image acquisition module shoots that people except people in the white list enter the house under the condition that the house of the owner is unoccupied, an alarm is given.
The present invention also provides an integrated anti-theft management system based on artificial intelligence, as shown in fig. 2, including:
a first obtaining module 1, configured to obtain a first image of a person entering a cell;
the first determining module 2 is used for determining suspicious people based on the first image and a preset portrait library;
the second acquisition module 3 is used for tracking and acquiring a second image of the suspicious person in the cell;
and the second determining module 4 is used for determining the theft behavior and outputting an alarm based on the second image and a preset neural network model.
In one embodiment, the first acquiring module 1 acquires a first image of a person entering a cell, and performs the following operations:
acquiring a first image of a person entering a cell by a law enforcement instrument worn by a security guard at a cell doorway;
and/or the presence of a gas in the gas,
a first image of a person entering a cell is acquired through a first image acquisition device arranged beside an entrance and an exit of the cell.
In one embodiment, the first determination module 2 determines the suspicious person based on the first image and a preset portrait base, and performs the following operations:
extracting the human body contour of the first image;
when the number of the extracted human body contours is one, matching the images corresponding to the human body contours with the human images in the human image library to determine whether the people corresponding to the human body contours are suspicious people;
when the number of the extracted human body contours is more than one, determining the relative positions of the human body contours and whether an overlapping area exists or not;
extracting features of the relative position and the overlapping area based on a preset first feature extraction template to obtain a plurality of feature values;
constructing a relation description vector between the human body contours based on the characteristic values;
acquiring a preset relation judgment library;
determining the relationship between the human body contours based on the relationship judgment library and the relationship description vector;
grouping the extracted human body contours based on the relationship between the human body contours to obtain a plurality of groups;
extracting any human body contour of each group according to a preset extraction rule, matching an image corresponding to the human body contour with each human image in a human image library, and determining whether a person corresponding to the human body contour is a suspicious person or not, wherein when the person corresponding to any human body contour in the same group is not the suspicious person, other persons in the same group are not the suspicious person;
wherein, the portrait base is constructed by the following steps:
acquiring a third image of each owner in the cell;
extracting a portrait of the third image;
and/or the presence of a gas in the atmosphere,
acquiring historical records of cell access;
analyzing the historical records, and determining fourth images of all people of which the times of entering and exiting the cell in a preset time period reach a preset time threshold;
and extracting the portrait of the fourth image.
In one embodiment, the second acquiring module 3 tracks and acquires a second image of the suspicious person in the cell, and performs the following operations:
constructing a virtual map of monitoring facilities of a cell;
acquiring the position, the moving direction and the moving speed of a suspicious person;
mapping the suspicious person to a virtual map based on the location of the suspicious person;
determining a second image acquisition device arranged in the cell for tracking shooting and a corresponding shooting start time for determining a second image based on the moving direction, the moving speed and the virtual map;
acquiring a second image of the suspicious person through second image acquisition equipment;
wherein determining a second image capturing device disposed within the cell for tracking shooting and a corresponding shooting start time for determining the second image based on the moving direction, the moving speed, and the virtual map comprises:
determining a first direction vector based on the moving direction and the position of the suspicious person in the virtual map;
determining a plurality of second direction vectors based on the position of the suspicious person in the virtual map and the setting position of each second image acquisition device;
respectively calculating the included angle between each second direction vector and the first direction vector;
extracting second image acquisition equipment corresponding to a second direction vector with an included angle smaller than a preset threshold value;
determining the distance between each extracted second image acquisition device and the position of the suspicious person;
and the second image acquisition device with the minimum distance is used as the second image acquisition device for tracking and shooting.
In one embodiment, the second determining module 4 determines the theft and outputs an alarm based on the second image and a preset neural network model, and performs the following operations:
extracting the features of the second image based on a preset second feature extraction template to obtain a plurality of behavior feature values;
inputting a plurality of behavior characteristic values into a neural network model, and determining a behavior analysis result;
and outputting an alarm when the behavior analysis result is the theft behavior.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.

Claims (10)

1. An integrated anti-theft management method based on artificial intelligence is characterized by comprising the following steps:
acquiring a first image of a person entering a cell;
determining suspicious people based on the first image and a preset portrait base;
tracking and acquiring a second image of the suspicious person in the cell;
and determining the theft behavior and outputting an alarm based on the second image and a preset neural network model.
2. The artificial intelligence based comprehensive anti-theft management method according to claim 1, wherein said obtaining a first image of a person entering a cell comprises:
acquiring the first image of a person entering a cell by a law enforcement instrument worn by security guards at a cell doorway;
and/or the presence of a gas in the gas,
the method comprises the steps that a first image of a person entering a community is acquired through first image acquisition equipment arranged beside an entrance and an exit of the community.
3. The integrated antitheft management method based on artificial intelligence of claim 1, wherein said determining suspicious people based on said first image and a preset portrait base comprises:
extracting a human body contour from the first image;
when the number of the extracted human body contours is one, matching the images corresponding to the human body contours with the individual human images in the human image library to determine whether the people corresponding to the human body contours are the suspicious people;
when the number of the extracted human body contours is more than one, determining the relative positions of the human body contours and whether an overlapping area exists;
extracting features of the relative position and the overlapping area based on a preset first feature extraction template to obtain a plurality of feature values;
constructing a relation description vector between the human body contours based on the characteristic values;
acquiring a preset relation judgment library;
determining the relationship between the human body contours based on the relationship judgment library and the relationship description vector;
grouping the extracted human body contours based on the relationship between the human body contours to obtain a plurality of groups;
extracting any human body contour of each group according to a preset extraction rule, matching an image corresponding to the human body contour with each human figure in a human figure library, and determining whether a person corresponding to the human body contour is a suspicious person or not, wherein when the person corresponding to any human body contour in the same group is not the suspicious person, other persons in the same group are not the suspicious person;
wherein the portrait library is constructed by:
acquiring a third image of each owner in the cell;
extracting a portrait of the third image;
and/or the presence of a gas in the gas,
acquiring a history record of cell access;
analyzing the historical records, and determining fourth images of all persons of which the times of entering and exiting the community in a preset time period reach a preset time threshold;
and extracting the portrait of the fourth image.
4. The integrated artificial intelligence based theft management method according to claim 1, wherein said tracking acquiring a second image of said suspicious person in a cell comprises:
constructing a virtual map of a monitoring facility of a cell;
acquiring the position, the moving direction and the moving speed of the suspicious personnel;
mapping the suspicious person to the virtual map based on the location of the suspicious person;
determining a second image acquisition device arranged in the cell for tracking shooting and a corresponding shooting start time for determining a second image based on the moving direction, the moving speed and the virtual map;
acquiring a second image of the suspicious person through the second image acquisition device;
wherein determining a second image capturing device disposed within the cell for tracking shooting and a corresponding shooting start time for determining a second image based on the moving direction, the moving speed, and the virtual map comprises:
determining a first direction vector based on the direction of movement and the location of the suspect in the virtual map;
determining a plurality of second direction vectors based on the position of the suspicious person in the virtual map and the setting position of each second image acquisition device;
respectively calculating included angles between the second direction vectors and the first direction vectors;
extracting the second image acquisition equipment corresponding to the second direction vector with the included angle smaller than a preset threshold value;
determining the distance between each extracted second image acquisition device and the position of the suspicious person;
and taking the second image acquisition equipment with the minimum distance as the second image acquisition equipment for tracking and shooting.
5. The integrated antitheft management method based on artificial intelligence of claim 1, wherein said determining theft and outputting an alarm based on said second image and a preset neural network model comprises:
extracting features of the second image based on a preset second feature extraction template to obtain a plurality of behavior feature values;
inputting a plurality of behavior characteristic values into the neural network model, and determining a behavior analysis result;
and outputting an alarm when the behavior analysis result is a theft behavior.
6. An integrated anti-theft management system based on artificial intelligence, comprising:
the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring a first image of a person entering a cell;
the first determination module is used for determining suspicious people based on the first image and a preset portrait library;
the second acquisition module is used for tracking and acquiring a second image of the suspicious person in the cell;
and the second determination module is used for determining the theft behavior and outputting an alarm based on the second image and a preset neural network model.
7. The integrated artificial intelligence based theft management system according to claim 6, wherein said first obtaining module obtains a first image of a person entering a cell by:
acquiring the first image of a person entering a cell by a law enforcement instrument worn by security guards at a cell doorway;
and/or the presence of a gas in the gas,
the method comprises the steps that a first image of a person entering a community is acquired through first image acquisition equipment arranged beside an entrance and an exit of the community.
8. The integrated artificial intelligence based theft management system according to claim 6, wherein said first determining module determines suspicious persons based on said first image and a preset portrait base, and performs the following operations:
extracting the human body contour of the first image;
when the number of the extracted human body contours is one, matching the image corresponding to the human body contours with each human figure in the human figure library to determine whether the person corresponding to the human body contours is the suspicious person;
when the number of the extracted human body contours is more than one, determining the relative positions of the human body contours and whether an overlapping area exists;
extracting features of the relative position and the overlapping area based on a preset first feature extraction template to obtain a plurality of feature values;
constructing a relation description vector between the human body contours based on the characteristic values;
acquiring a preset relation judgment library;
determining the relationship between the human body contours based on the relationship judgment library and the relationship description vector;
grouping the extracted human body contours based on the relationship between the human body contours to obtain a plurality of groups;
extracting any human body contour of each group according to a preset extraction rule, matching an image corresponding to the human body contour with each human figure in a human figure library, and determining whether a person corresponding to the human body contour is a suspicious person or not, wherein when the person corresponding to any human body contour in the same group is not the suspicious person, other persons in the same group are not the suspicious person;
wherein the portrait library is constructed by:
acquiring a third image of each owner in the cell;
extracting a portrait of the third image;
and/or the presence of a gas in the gas,
acquiring historical records of cell access;
analyzing the historical records, and determining fourth images of all persons of which the times of entering and exiting the community in a preset time period reach a preset time threshold;
and extracting the portrait of the fourth image.
9. The artificial intelligence based integrated antitheft management system of claim 6 wherein the second acquisition module tracks acquisition of a second image of the suspect person within the cell by performing the following operations:
constructing a virtual map of a monitoring facility of a cell;
acquiring the position, the moving direction and the moving speed of the suspicious personnel;
mapping the suspicious person to the virtual map based on the location of the suspicious person;
determining a second image acquisition device arranged in the cell for tracking shooting and a corresponding shooting start time for determining a second image based on the moving direction, the moving speed and the virtual map;
acquiring a second image of the suspicious person through the second image acquisition device;
wherein determining a second image capturing device disposed within the cell for tracking shooting and a corresponding shooting start time for determining a second image based on the moving direction, moving speed, and the virtual map comprises:
determining a first direction vector based on the direction of movement and the location of the suspect in the virtual map;
determining a plurality of second direction vectors based on the position of the suspicious person in the virtual map and the setting position of each second image acquisition device;
respectively calculating included angles between the second direction vectors and the first direction vectors;
extracting the second image acquisition equipment corresponding to the second direction vector with the included angle smaller than a preset threshold value;
determining the distance between each extracted second image acquisition device and the position of the suspicious person;
and taking the second image acquisition equipment with the minimum distance as the second image acquisition equipment for tracking and shooting.
10. The integrated artificial intelligence based theft management system according to claim 6, wherein said second determining module determines theft and outputs an alarm based on said second image and a preset neural network model, and performs the following operations:
performing feature extraction on the second image based on a preset second feature extraction template to obtain a plurality of behavior feature values;
inputting a plurality of behavior characteristic values into the neural network model, and determining a behavior analysis result;
and outputting an alarm when the behavior analysis result is a theft behavior.
CN202210793697.7A 2022-07-05 2022-07-05 Comprehensive anti-theft management method and system based on artificial intelligence Active CN115294515B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210793697.7A CN115294515B (en) 2022-07-05 2022-07-05 Comprehensive anti-theft management method and system based on artificial intelligence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210793697.7A CN115294515B (en) 2022-07-05 2022-07-05 Comprehensive anti-theft management method and system based on artificial intelligence

Publications (2)

Publication Number Publication Date
CN115294515A true CN115294515A (en) 2022-11-04
CN115294515B CN115294515B (en) 2023-06-13

Family

ID=83821865

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210793697.7A Active CN115294515B (en) 2022-07-05 2022-07-05 Comprehensive anti-theft management method and system based on artificial intelligence

Country Status (1)

Country Link
CN (1) CN115294515B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3252714A1 (en) * 2016-06-03 2017-12-06 Univrses AB Camera selection in positional tracking
CN107592498A (en) * 2017-08-28 2018-01-16 深圳市盛路物联通讯技术有限公司 A kind of cell management method and relevant device based on intelligent video camera head
CN108009466A (en) * 2016-10-28 2018-05-08 北京旷视科技有限公司 Pedestrian detection method and device
CN109473168A (en) * 2018-10-09 2019-03-15 五邑大学 A kind of medical image robot and its control, medical image recognition methods
CN109614875A (en) * 2018-11-16 2019-04-12 合肥未来计算机技术开发有限公司 A kind of intelligent safety prevention warning system based on sports rule
CN110674761A (en) * 2019-09-27 2020-01-10 三星电子（中国）研发中心 Regional behavior early warning method and system
CN111860118A (en) * 2020-06-03 2020-10-30 安徽碧耕软件有限公司 Human behavior analysis method based on artificial intelligence
WO2021213141A1 (en) * 2020-04-20 2021-10-28 华为技术有限公司 Data processing method and device, and apparatus

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3252714A1 (en) * 2016-06-03 2017-12-06 Univrses AB Camera selection in positional tracking
CN108009466A (en) * 2016-10-28 2018-05-08 北京旷视科技有限公司 Pedestrian detection method and device
CN107592498A (en) * 2017-08-28 2018-01-16 深圳市盛路物联通讯技术有限公司 A kind of cell management method and relevant device based on intelligent video camera head
CN109473168A (en) * 2018-10-09 2019-03-15 五邑大学 A kind of medical image robot and its control, medical image recognition methods
CN109614875A (en) * 2018-11-16 2019-04-12 合肥未来计算机技术开发有限公司 A kind of intelligent safety prevention warning system based on sports rule
CN110674761A (en) * 2019-09-27 2020-01-10 三星电子（中国）研发中心 Regional behavior early warning method and system
WO2021213141A1 (en) * 2020-04-20 2021-10-28 华为技术有限公司 Data processing method and device, and apparatus
CN111860118A (en) * 2020-06-03 2020-10-30 安徽碧耕软件有限公司 Human behavior analysis method based on artificial intelligence

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
WAQAS AHMED等: "Robust Suspicious Action Recognition Approach Using Pose Descriptor", 《MATHEMATICAL PROBLEMS IN ENGINEERING》 *

Also Published As

Publication number Publication date
CN115294515B (en) 2023-06-13

Similar Documents

Publication Publication Date Title
CN110491004B (en) Resident community personnel safety management system and method
CN110191424B (en) Specific suspect track generation method and apparatus
CN109658554B (en) Intelligent residential district security protection system based on big data
CN112233300A (en) Community passing epidemic prevention monitoring joint defense system and method based on artificial intelligence
US9911294B2 (en) Warning system and method using spatio-temporal situation data
CN110674761B (en) Regional behavior early warning method and system
WO2018180588A1 (en) Facial image matching system and facial image search system
CN110766895A (en) Intelligent community abnormity alarm system and method based on target trajectory analysis
CN108256459A (en) Library algorithm is built in detector gate recognition of face and face based on multiple-camera fusion automatically
WO2004042673A2 (en) Automatic, real time and complete identification of vehicles
CN103839373A (en) Sudden abnormal event intelligent identification alarm device and system
CN108710827B (en) A kind of micro- police service inspection in community and information automatic analysis system and method
KR102012672B1 (en) Anti-crime system and method using face recognition based people feature recognition
CN106951846A (en) A kind of face 3D models typing and recognition methods and device
CN112183162A (en) Face automatic registration and recognition system and method in monitoring scene
CN114548590B (en) Internet of things-based engineering finishing intelligent management platform and method
JP5349080B2 (en) Admission management system, admission management device, and admission management method
KR102142315B1 (en) ATM security system based on image analyses and the method thereof
CN112818854B (en) All-weather video person searching method applied to campus security
CN115294515A (en) Artificial intelligence-based comprehensive anti-theft management method and system
KR20180137155A (en) System for smart cpted based reinforcing approach control and method by using the same
CN116778657A (en) Method and system for intelligently identifying intrusion behavior
CN113554682B (en) Target tracking-based safety helmet detection method
CN116052035A (en) Power plant personnel perimeter intrusion detection method based on convolutional neural network
CN114926930A (en) Intelligent building monitoring system with multiple identity verifications and method thereof

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
GR01 Patent grant
GR01 Patent grant