CN113239742A - System and method for monitoring abnormal gathering behavior of people in community - Google Patents

System and method for monitoring abnormal gathering behavior of people in community Download PDF

Info

Publication number
CN113239742A
CN113239742A CN202110442521.2A CN202110442521A CN113239742A CN 113239742 A CN113239742 A CN 113239742A CN 202110442521 A CN202110442521 A CN 202110442521A CN 113239742 A CN113239742 A CN 113239742A
Authority
CN
China
Prior art keywords
abnormal
gathering
people
electricity
water consumption
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
Application number
CN202110442521.2A
Other languages
Chinese (zh)
Inventor
张天骏
陈可鑫
张震
张安琪
宋超
陈云飞
王晓杰
晋志华
王琪
吴云鹏
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.)
Zhengzhou University
Original Assignee
Zhengzhou University
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 Zhengzhou University filed Critical Zhengzhou University
Priority to CN202110442521.2A priority Critical patent/CN113239742A/en
Publication of CN113239742A publication Critical patent/CN113239742A/en
Pending legal-status Critical Current

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
    • G06V40/10—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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
    • 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
    • G06V20/49—Segmenting video sequences, i.e. computational techniques such as parsing or cutting the sequence, low-level clustering or determining units such as shots or scenes
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Mathematical Physics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Probability & Statistics with Applications (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Human Computer Interaction (AREA)
  • Alarm Systems (AREA)

Abstract

The invention belongs to the technical field of big data analysis, and particularly relates to a monitoring system and a monitoring method for abnormal gathering behavior of people in a community, wherein the system comprises a similar attribute abnormal gathering monitoring subsystem, a data processing subsystem and a data processing subsystem, wherein the similar attribute abnormal gathering monitoring subsystem is used for acquiring abnormal gathering information of similar attribute people; the electricity and water consumption abnormity detection subsystem is used for acquiring household information of electricity and water consumption abnormity; and the abnormal gathering household screening subsystem is respectively connected with the similar attribute person abnormal gathering monitoring subsystem and the electricity and water consumption abnormality detection subsystem, and screens out households which have the similar attribute person abnormal gathering behavior and have abnormal household electricity and water consumption in all households. The method and the system preliminarily screen the residential areas and buildings which are likely to have abnormal gathering of people with the same attribute, and perform secondary screening on the residents in the residential areas by combining the abnormal electricity and water consumption detection data of the residents, so that the monitoring intensity and accuracy are improved, and the human resource consumption is reduced.

Description

System and method for monitoring abnormal gathering behavior of people in community
Technical Field
The invention belongs to the technical field of big data analysis, and particularly relates to a system and a method for monitoring abnormal gathering behaviors of people in a community.
Background
In recent years, with the development of the fields of big data, cloud computing, artificial intelligence and the like, rapid growth of data brings severe challenges and precious opportunities to many industries, and big data has huge value, has important significance and significance in different fields of society, economy, scientific research and the like, and is deep into the lives of people to provide sufficient information for people to know and understand the physical world more thoroughly.
At present, a complete and reasonable monitoring system and method for abnormal gathering behaviors of people does not exist, the monitoring of the abnormal gathering behaviors of people mainly depends on reporting of people, and the problems of resource consumption, monitoring strength and accuracy are high. Therefore, how to design a system and a method for monitoring abnormal gathering behavior of people based on big data analysis is urgent.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention aims to provide a system and a method for monitoring abnormal gathering behaviors of people in a community, which can realize real-time monitoring and alarming on the abnormal gathering behaviors of the people and improve the monitoring strength and accuracy.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention provides a monitoring system for abnormal gathering behavior of people in a community, which comprises:
the system comprises a similar attribute personnel abnormal gathering monitoring subsystem, a similar attribute personnel abnormal gathering monitoring subsystem and a similar attribute personnel abnormal gathering monitoring subsystem, wherein the similar attribute personnel abnormal gathering monitoring subsystem is used for acquiring similar attribute personnel abnormal gathering information;
the electricity and water consumption abnormity detection subsystem is used for acquiring household information of electricity and water consumption abnormity;
and the abnormal gathering household screening subsystem is respectively connected with the similar attribute person abnormal gathering monitoring subsystem and the electricity consumption water abnormality detection subsystem, and screens out households which have the similar attribute person abnormal gathering behavior and abnormal household electricity consumption water consumption in all households.
Further, the system for monitoring the abnormal gathering of the persons with the same type of attributes comprises:
the high-definition cameras are installed at a community entrance guard and a building entrance guard and are used for shooting videos of people entering and exiting the community and the building, carrying out a face detection algorithm on the video stream, cutting the frame of picture if face information is found, and storing the face picture in a first database;
the data processing and analyzing module I is used for identifying the face information in the database I, acquiring personnel identity information and label information, storing an analysis result into the database, performing collision analysis on the analysis result and data in the public security service library, preliminarily judging that abnormal gathering behaviors of personnel are suspected to occur when the number of similar attribute personnel reaches a preset threshold value, and recording a screening result into an abnormal gathering database of the similar attribute personnel;
and the first display module is used for displaying the information of the people entering and exiting the community, the information of the people entering and exiting the building and the screening result on a visual interface.
Further, the face detection algorithm adopts a Centerface, and specifically includes:
the center face structure is as follows: the whole structure adopts a MobileNet V2 structure, MobileNet V2 carries out downsampling for 5 times, 3 upsampling layers are added on the last layer of MobileNet V2, the size of final output is carried out downsampling for 2 times, and the output dimension is 1/4 of the original image;
defining a target center point: the target center point is
Figure BDA0003035694800000021
Wherein (x)1,y1)(x2,y2) Coordinates corresponding to the upper left corner and the lower right corner of the target;
the loss function for face detection is as follows:
face classification loss function:
Figure BDA0003035694800000031
where alpha and beta are hyper-parameters of the focal loss,
Figure BDA0003035694800000032
indicating that a key point has been detected and,
Figure BDA0003035694800000033
denotes the background, YxycThe number of the gaussian kernels is represented,
Figure BDA0003035694800000034
w is the image width, H is the image height, C is the number of keypoint types;
face frame center point offset loss function:
Figure BDA0003035694800000035
Figure BDA0003035694800000036
wherein o isκWhich represents the predicted coordinates of the object to be predicted,
Figure BDA0003035694800000037
coordinates representing the GT box, N representing the number of keypoints in the image, xκAnd yκIs the x, y coordinates of the face center k, n represents the down-sampling factor;
face frame loss function:
Figure BDA0003035694800000038
wherein x1,x2,y1,y2The coordinates of the face frame are represented,
Figure BDA0003035694800000039
the loss in the longitudinal direction is shown,
Figure BDA00030356948000000310
represents the lateral loss, R represents the step size;
overall loss function: l ═ Lc+λoffLoff+λboxLboxWhere λ is the loss function coefficient.
Further, the algorithm for recognizing the face information in the first database adopts yolov5, and specifically includes:
dividing an input face picture into S multiplied by S grids, wherein each cell is responsible for detecting a face image with a central point in the grid, and each cell can predict B bounding boxes and confidence degrees of the bounding boxes, wherein the confidence degree is defined as
Figure BDA00030356948000000311
Pr (object) indicates the probable size of the bounding box containing the face image,
Figure BDA00030356948000000312
indicating the accuracy of the bounding box;
giving predicted C class probability values for each cell, and representing the probability Pr (class) that the cell is responsible for predicting the bounding box and the target of the cell belong to each classiI object), calculating confidence of each bounding box category
Figure BDA0003035694800000041
The confidence of the bounding box category characterizes the probability of the target in the bounding box belonging to each category and how well the bounding box matches the target.
Further, the electricity water abnormality detection subsystem includes:
the front-end water inductance sensing equipment is used for acquiring the electricity and water consumption of the household every hour and storing the data into a second database;
the second data processing and analyzing module is used for calculating the average electricity consumption and water consumption of the residents according to the electricity consumption and water consumption within one hour and the number of the resident population, and storing the average electricity consumption and water consumption in the second database; calculating the average electricity consumption of all residents in the community according to the electricity consumption of all residents in the community and the number of the residents in the community; when inquiring out the residents with the average electricity consumption of more than 1.5 times of the average electricity consumption of the residents and the residents with the average water consumption of more than 1.5 times of the average water consumption of the residents, marking, and storing the marking result into an abnormal electricity and water consumption database of the residents;
and the display module II is used for displaying the average electricity consumption of the residents, the average electricity consumption of the community, the average water consumption of the residents and the average water consumption of the community on a visual interface on the same day.
Further, the abnormal aggregated resident screening subsystem comprises:
the input port is respectively connected with the abnormal gathering database of the persons with the same attribute and the abnormal electricity and water consumption database of the residents;
the data processing and analyzing module III is used for jointly inquiring the data in the two databases, screening out the situations that the abnormal gathering behaviors of people with the same type of attributes exist in a certain building and the electricity and water consumption of a certain resident are abnormal, and storing the data into a suspected abnormal gathering behavior database of people;
and a display module III for displaying the data in the suspected person abnormal gathering behavior database on a visual interface.
The invention also provides a method for monitoring the abnormal gathering behavior of the people in the community, which comprises the following steps:
acquiring abnormal gathering information of persons with similar attributes;
acquiring household information of abnormal electricity and water;
and screening out residents with abnormal gathering behaviors of people with the same attribute and abnormal household electricity and water consumption.
Further, the acquiring of the abnormal gathering information of the persons with the same type of attributes comprises the following steps:
the high-definition camera shoots videos of people entering and exiting a community and building, a face detection method is carried out on a video stream, if face information is found, the frame of picture is cut, the face picture is stored in a first database and is sent to a first data processing and analyzing module;
the data processing and analyzing module identifies the face information and stores the screening result into an abnormal gathering database of the similar attribute personnel;
and displaying the screening result on a visual interface.
Further, the acquiring of the household information with abnormal electricity and water consumption comprises the following steps:
the front-end water inductance sensing equipment acquires the electricity consumption and water consumption of the household every one hour, stores the data into a second database and sends the data to a second data processing and analyzing module;
the data processing and analyzing module processes and analyzes the data and stores the result into a household abnormal water consumption database;
and displaying the processing result on a visual interface.
Further, the screening out the residents with the abnormal gathering of the persons with the similar attributes and the abnormal household electricity and water consumption comprises the following steps:
the data processing and analyzing module is used for inquiring data in the abnormal gathering database of the persons with the same attribute and the abnormal electricity and water consumption database of the residents in a triple-combination manner, screening out the conditions that the persons with the same attribute abnormally gather in a certain building and the residents have abnormal electricity and water consumption, and storing the data in the abnormal gathering behavior database of the suspected persons;
and displaying the data in the database of the abnormal gathering behavior of the suspected personnel on a visual interface.
Compared with the prior art, the invention has the following advantages:
the monitoring system for the abnormal gathering behavior of the people in the community comprises a similar attribute abnormal gathering monitoring subsystem, an electricity and water consumption abnormal detecting subsystem and an abnormal gathering resident screening subsystem, can acquire the identity information of people entering and exiting a community and a building under the unattended condition, preliminarily screens the community and the building which are likely to have the abnormal gathering of the similar attribute people, combines the electricity and water consumption abnormal detection data of the residents, and carries out secondary screening on the residents in the community.
The face detection algorithm of the invention adopts the center, can be used when the flow of people in a community is large during work and work, has higher accuracy and working efficiency, and avoids face information missing detection; the face recognition algorithm adopts yolov5, and the accuracy and efficiency of face recognition can be ensured.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly described below, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise.
FIG. 1 is a block diagram of a system for monitoring abnormal gathering behavior of people in a community according to an embodiment of the present invention;
FIG. 2 is a diagram of a Centerface algorithm model according to an embodiment of the invention;
fig. 3 is a schematic flow chart of a method for monitoring abnormal people gathering behavior in a community according to an embodiment of the present invention.
Detailed Description
In order to make those skilled in the art better understand the technical solution of the present invention, the technical solution in the embodiment of the present invention will be clearly and completely described below with reference to the drawings in the embodiment of the present invention, and it is obvious that the described embodiment is only a part of the embodiment of the present invention, and not all embodiments.
As shown in fig. 1, the monitoring system for abnormal people gathering behavior in a community of this embodiment includes a same-category abnormal people gathering monitoring subsystem 11, an electricity and water consumption abnormality detecting subsystem 12, and an abnormal gathering resident screening subsystem 13; the similar attribute person abnormal gathering monitoring subsystem is used for acquiring similar attribute person abnormal gathering information; the electricity and water consumption abnormity detection subsystem is used for acquiring resident information of electricity and water consumption abnormity; the abnormal gathering resident screening subsystem is respectively connected with the similar attribute person abnormal gathering monitoring subsystem and the electricity and water consumption abnormality detecting subsystem, and screens out residents with the similar attribute person abnormal gathering behaviors and the household electricity and water consumption abnormality in all residents.
The system comprises a monitoring subsystem for monitoring the abnormal gathering of the similar attribute personnel, a high-definition camera, a data processing and analyzing module I and a display module I.
The high-definition cameras are installed at a community entrance guard and a building entrance guard, videos of people entering and exiting a community and building are mainly shot, face detection algorithm is carried out on the portrait videos shot by the equipment, if face information is found, the frame of picture is cut, and the face picture is stored in a first database.
The data processing and analyzing module I is used for identifying and analyzing the face information in the database I, acquiring personnel identity information and label information, storing an analysis result into the database I, performing collision analysis on the database I and data in a public security service library, inquiring whether the personnel in and out are persons with certain attributes, preliminarily judging the behavior of suspected abnormal gathering of the personnel when the number of the persons with the same attributes reaches a preset threshold value, and recording a screening result into the abnormal gathering database of the persons with the same attributes.
The first display module is mainly used for displaying information of people who enter and exit from a community, information of people who enter and exit from a building and a screening result on a visual interface, and therefore system management personnel can conveniently check the information.
Because the people flow of the community is large during the commuting period, in order to improve the working efficiency of the system, the accuracy and the efficiency must be considered in the aspect of selecting the face detection algorithm, the face detection algorithm in the embodiment adopts a center face, and the structure is shown in fig. 2.
The center face structure is as follows: the whole structure adopts a MobileNet V2 structure, MobileNet V2 carries out downsampling for 5 times, 3 upsampling layers are added on the last layer of MobileNet V2, the size of final output is carried out downsampling for 2 times, and the output dimension is 1/4 of the original image.
Defining a target center point: the target center point is
Figure BDA0003035694800000081
Wherein (x)1,y1)(x2,y2) Coordinates corresponding to the top left and bottom right corners of the target. In the center, whether the coordinates of the center point are in the image corresponding to the feature map is calculated,
Figure BDA0003035694800000082
r is the corresponding step length, the feature map where the central point is 1, and for the label, points are scattered outwards in a Gaussian mode.
In supervised machine learning algorithms, it is desirable to minimize the error per training sample during the learning process, and this error comes from the loss function, which is the following for face detection:
face classification loss function:
Figure BDA0003035694800000083
where α and β are hyper-parameters of the focal loss, specifying α -2, β -4,
Figure BDA0003035694800000084
indicating that a key point has been detected and,
Figure BDA0003035694800000085
denotes the background, YxycThe number of the gaussian kernels is represented,
Figure BDA0003035694800000086
w is the image width, H is the image height, and C is the number of keypoint types.
Face frame center point offset loss function:
Figure BDA0003035694800000087
Figure BDA0003035694800000088
wherein o isκWhich represents the predicted coordinates of the object to be predicted,
Figure BDA0003035694800000091
coordinates representing the GT box, N representing the number of keypoints in the image, xκAnd yκIs the x, y coordinate of the face center k, n represents the down-sampling factor.
Face frame loss function:
Figure BDA0003035694800000092
wherein x1,x2,y1,y2The coordinates of the face frame are represented,
Figure BDA0003035694800000093
the loss in the longitudinal direction is shown,
Figure BDA0003035694800000094
represents the lateral loss and R represents the step size.
Overall loss function: l ═ Lc+λoffLoff+λboxLboxWhere λ is the loss function coefficient. After the loss function is determined, in the process of training data, the size of the loss function coefficient can be properly adjusted according to the difference between the prediction result and the actual result, so that the loss function is small as much as possible, and the accuracy rate of the algorithm is high.
In order to ensure the accuracy and efficiency of face recognition, the adopted algorithm is yolov5, which specifically comprises the following steps:
the CNN network of Yolo divides an input picture into S × S meshes, and then each cell is negativeIt is important to detect those face images whose center points fall within the grid. The confidence level of each cell, which includes two aspects, i.e., the probability that the bounding box contains an object, and the accuracy of the bounding box, which is denoted as pr (object), when the bounding box is background (i.e., does not contain an object), pr (object) is 0, and when the bounding box contains an object, pr (object) is 1; the accuracy of the bounding box can be characterized by the IOU of the predicted and actual boxes, noted
Figure BDA0003035694800000095
Confidence may therefore be defined as
Figure BDA0003035694800000096
The size and position of the bounding box can be characterized by 4 values: (x, y, w, h), where (x, y) is the center coordinates of the bounding box, and w and h are the width and height of the bounding box. The predicted value (x, y) of the center coordinate is an offset value with respect to the coordinate point at the upper left corner of each cell, and the unit is with respect to the cell size; the w and h predictors of the bounding box are the ratio of width to height relative to the whole picture, so theoretically the size of 4 elements should be in the 0,1 range. Thus, the prediction value of each bounding box actually contains 5 elements: (x, y, w, h, c), where the first 4 characterize the size and position of the bounding box and the last value is the confidence.
For each cell, a probability value of predicting C classes is also given, which characterizes the probability Pr (class) of the bounding box and its target belonging to each class for which the cell is responsible for predictingTI object). We can calculate the confidence for each bounding box class:
Figure BDA0003035694800000101
Figure BDA0003035694800000102
the confidence of the bounding box category characterizes the probability size and the edge of the target in the bounding box belonging to each categoryThe bounding box matches the quality of the target.
The power utilization water abnormity detection subsystem comprises front-end water inductance sensing equipment, a second data processing and analyzing module and a second display module.
The front-end water inductance sensing equipment acquires the electricity and water consumption of the household every hour (24 times a day), stores the data in the second database and sends the data to the second data processing and analyzing module through the network.
After receiving the data sent by the front-end water-electricity sensing equipment, the data processing and analyzing module II takes the household as a unit, inquires the number of the daily living population of the household from the database II, calculates the average electricity consumption and water consumption of the household according to the electricity consumption and water consumption in one hour and the daily living population of the household, and stores the average electricity consumption and water consumption in the database II; and calculating the average electricity consumption of the community according to the electricity consumption of all residents in the community and the number of the community permanent population, and storing the average electricity consumption in the second database. When inquiring out the residents with the average electricity consumption of the residents more than 1.5 times larger than the average electricity consumption of the residents in the community and the average water consumption of the residents more than 1.5 times larger than the average water consumption of the residents in the community, marking the residents, and storing the marking result into an abnormal electricity and water consumption database of the residents so as to be conveniently screened for the second time by the abnormal gathering monitoring subsystem of the similar attribute personnel.
And the second display module displays the average electricity consumption of residents, the average electricity consumption of community people, the average water consumption of residents and the average water consumption of community people on a visual interface on the same day, preferably, the display module can display by adopting a line graph.
The abnormal aggregated resident screening subsystem comprises an input port, a third data processing and analyzing module and a third display module.
The input port is respectively connected with the similar attribute personnel abnormal gathering database and the household electricity consumption abnormal quantity database.
And the data processing and analyzing module III jointly inquires the data in the two databases, screens out the situations that the people with the same type of attributes are abnormally gathered in a certain building and the electricity and water consumption of a certain household are abnormal, and stores the data in the suspected people abnormal gathering behavior database.
And the display module III displays the data in the suspected person abnormal gathering behavior database on a visual interface.
As shown in fig. 3, the embodiment further provides a method for monitoring abnormal gathering behavior of people in a community, which includes the following steps:
and step S31, acquiring abnormal aggregation information of persons with the same type of attributes, judging whether abnormal aggregation behaviors of the persons with the same type of attributes exist in the cell and the building, and primarily screening the abnormal aggregation behaviors of the suspected persons. The method specifically comprises the following steps:
step S311, the high-definition camera shoots videos of people entering and leaving the residential area and the building in real time, a face detection algorithm is carried out on the video stream, if face information is found, the frame of picture is intercepted, the face picture is stored in a first database and stored, and the face picture is sent to a first data processing and analyzing module.
Step S312, the data processing and analyzing module identifies the face information to obtain the information of the person, such as the name, the identification number, whether the person participates in illegal events, and the like, performs collision analysis with the data of the public security service library to judge whether the target person is suspected abnormal gathering person, and then stores the screening result into the abnormal gathering database of the similar attribute person, which can provide convenience for subsequent big data analysis.
And step S313, displaying the screening result on a visual interface.
By the monitoring method for the abnormal gathering behavior of the similar attribute personnel, data analysis can be carried out on the people who enter and exit the community and the building within 24 hours, whether a large number of similar attribute personnel are abnormally gathered or not is judged, the situation of entering and exiting the building is accurately achieved, and the reliability is higher.
And step S32, acquiring the information of residents with abnormal electricity consumption, judging whether residents with household water consumption greatly exceeding the average electricity consumption of the community exist, and performing secondary screening on suspected abnormal gathering behaviors of people. The method specifically comprises the following steps:
and S321, acquiring the electricity consumption of the household every other hour (24 times a day) by the front-end water inductance sensing equipment, storing the data in a second database, and sending the data to a second data processing and analyzing module.
And S322, processing and analyzing the two pairs of data by the data processing and analyzing module, screening out residents with the average electricity consumption of the residents more than 1.5 times that of the community and the average water consumption of the residents more than 1.5 times that of the community, marking, and storing the result in an abnormal electricity and water consumption database of the residents.
And step S323, displaying the processing result on a visual interface, so that system management personnel can visually find problems and more clearly find out which resident is likely to have abnormal personnel gathering.
By the monitoring method for the household electricity consumption abnormity, the data of the household electricity consumption can be obtained and analyzed, and the household with abnormal household electricity consumption in the community can be screened out, so that the reliability is higher.
And step S33, screening out residents with abnormal gathering behaviors of similar attribute personnel and abnormal household electricity consumption water consumption among all the residents.
And step S331, the data processing and analyzing module is used for inquiring data in the similar attribute abnormal personnel gathering database and the household abnormal electricity and water consumption database in a triple mode, screening out the conditions that similar attribute abnormal gathering behaviors exist in a certain building and electricity and water consumption of a certain household exist, and storing the data in the suspected abnormal gathering behavior database.
And S332, displaying the data in the suspected person abnormal gathering behavior database on a visual interface, and reminding a system manager in a short message reminding manner.
Unless defined otherwise, technical or scientific terms used herein shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention belongs. The use of the terms "a" or "an" and the like in the description and in the claims of this application do not necessarily denote a limitation of quantity. The word "comprising" or "comprises", and the like, means that the element or item listed before the word covers the element or item listed after the word and its equivalent, but does not exclude other elements or items. The terms "connected" or "coupled" and the like are not restricted to physical or mechanical connections, but may include electrical connections, whether direct or indirect.
Finally, it is to be noted that: the above description is only a preferred embodiment of the present invention, and is only used to illustrate the technical solutions of the present invention, and not to limit the protection scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.

Claims (10)

1. An abnormal gathering behavior monitoring system for people in a community, comprising:
the system comprises a similar attribute personnel abnormal gathering monitoring subsystem, a similar attribute personnel abnormal gathering monitoring subsystem and a similar attribute personnel abnormal gathering monitoring subsystem, wherein the similar attribute personnel abnormal gathering monitoring subsystem is used for acquiring similar attribute personnel abnormal gathering information;
the electricity and water consumption abnormity detection subsystem is used for acquiring household information of electricity and water consumption abnormity;
and the abnormal gathering household screening subsystem is respectively connected with the similar attribute person abnormal gathering monitoring subsystem and the electricity and water consumption abnormality detection subsystem, and screens out households which have the similar attribute person abnormal gathering behavior and have abnormal household electricity and water consumption in all households.
2. The system for monitoring people abnormal gathering behavior in community according to claim 1, wherein the system for monitoring people abnormal gathering with similar attributes comprises:
the high-definition cameras are installed at a community entrance guard and a building entrance guard and are used for shooting videos of people entering and exiting the community and the building, carrying out a face detection algorithm on the video stream, cutting the frame of picture if face information is found, and storing the face picture in a first database;
the data processing and analyzing module I is used for identifying the face information in the database I, acquiring personnel identity information and label information, storing an analysis result into the database, performing collision analysis on the analysis result and data in the public security service library, preliminarily judging that abnormal gathering behaviors of personnel are suspected to occur when the number of similar attribute personnel reaches a preset threshold value, and recording a screening result into an abnormal gathering database of the similar attribute personnel;
and the first display module is used for displaying the information of the people entering and exiting the residential area, the information of the people entering and exiting the building and the screening result on a visual interface.
3. The system for monitoring abnormal gathering behavior of people in community according to claim 2, wherein the face detection algorithm employs a Centerface, and specifically comprises:
the center face structure is as follows: the whole structure adopts a MobileNet V2 structure, MobileNet V2 carries out downsampling for 5 times, 3 upsampling layers are added on the last layer of MobileNet V2, downsampling is carried out for 2 times according to the final output size, and the output dimension is 1/4 of the original image;
defining a target center point: the target center point is
Figure FDA0003035694790000021
Wherein (x)1，y1)(x2，y2) Coordinates corresponding to the upper left corner and the lower right corner of the target;
the loss function for face detection is as follows:
face classification loss function:
Figure FDA0003035694790000022
where alpha and beta are hyper-parameters of the focal loss,
Figure FDA0003035694790000023
indicating that a key point has been detected and,
Figure FDA0003035694790000024
denotes the background, YxycThe number of the gaussian kernels is represented,
Figure FDA0003035694790000025
w is the image width, H is the image height, C is the number of keypoint types;
face frame center point offset loss function:
Figure FDA0003035694790000026
Figure FDA0003035694790000027
wherein o isκWhich represents the predicted coordinates of the object to be predicted,
Figure FDA0003035694790000028
coordinates representing the GT box, N representing the number of keypoints in the image, xκAnd yκIs the x, y coordinates of the face center k, n represents the down-sampling factor;
face frame loss function:
Figure FDA0003035694790000029
wherein x1，x2，y1，y2The coordinates of the face frame are represented,
Figure FDA00030356947900000210
the loss in the longitudinal direction is shown,
Figure FDA00030356947900000211
represents the lateral loss, R represents the step size;
overall loss function: l ═ Lc+λoffLoff+λboxLboxWhere λ is the loss function coefficient.
4. The system for monitoring abnormal people gathering behavior in community according to claim 2, wherein the algorithm for recognizing the face information in the first database is yolov5, and specifically comprises:
dividing the input human face picture into S multiplied by S grids, each cell is responsible for detecting the human face image with the central point in the grid, and each cell can predict B bounding boxes and the confidence of the bounding boxesDegree, wherein the confidence level is defined as
Figure FDA0003035694790000031
Pr (object) represents the size of the probability that the bounding box contains a face image,
Figure FDA0003035694790000032
indicating the accuracy of the bounding box;
giving predicted C class probability values for each cell, and representing the probability Pr (class) that the cell is responsible for predicting the bounding box and the target of the cell belong to each classiI object), calculating confidence of each bounding box category
Figure FDA0003035694790000033
The confidence of the bounding box category characterizes the probability of the target in the bounding box belonging to each category and the quality of the bounding box matching the target.
5. The system for monitoring abnormal people gathering behavior in community according to claim 2, wherein the power utilization water abnormality detection subsystem comprises:
the front-end water inductance sensing equipment is used for acquiring the electricity and water consumption of the household every hour and storing the data into a second database;
the second data processing and analyzing module is used for calculating the average electricity consumption and water consumption of the residents according to the electricity consumption and water consumption within one hour and the number of the resident population, and storing the average electricity consumption and water consumption in the second database; calculating the average electricity consumption of all residents in the community according to the electricity consumption and water consumption of all residents in the community and the number of the ordinary living population of the community; when inquiring out the residents with the average electricity consumption of more than 1.5 times of the average electricity consumption of the residents and the residents with the average water consumption of more than 1.5 times of the average water consumption of the residents, marking, and storing the marking result into an abnormal electricity and water consumption database of the residents;
and the display module II is used for displaying the average electricity consumption of the residents, the average electricity consumption of the community, the average water consumption of the residents and the average water consumption of the community on a visual interface on the same day.
6. The system for monitoring people abnormal gathering behavior in community as claimed in claim 5, wherein the abnormal gathering resident screening subsystem comprises:
the input port is respectively connected with the abnormal gathering database of the persons with the same attribute and the abnormal electricity and water consumption database of the residents;
the data processing and analyzing module III is used for jointly inquiring the data in the two databases, screening out the situations that the abnormal gathering behaviors of people with the same type of attributes exist in a certain building and the electricity and water consumption of a certain resident are abnormal, and storing the data into a suspected abnormal gathering behavior database of people;
and a display module III for displaying the data in the suspected person abnormal gathering behavior database on a visual interface.
7. A method for monitoring abnormal gathering behaviors of people in a community is characterized by comprising the following steps:
acquiring abnormal gathering information of persons with similar attributes;
acquiring household information of abnormal electricity and water;
and screening out the residents with abnormal gathering behaviors of people with the same attribute and abnormal household electricity and water consumption.
8. The method for monitoring abnormal gathering behavior of people in community according to claim 7, wherein the step of obtaining the abnormal gathering information of people with similar attributes comprises the following steps:
the high-definition camera shoots videos of people entering and exiting a community and building, a face detection algorithm is carried out on the video stream, if face information is found, the frame of picture is cut, the face picture is stored in a first database and is sent to a first data processing and analyzing module;
the data processing and analyzing module identifies the face information and stores the screening result into an abnormal gathering database of the similar attribute personnel;
and displaying the screening result on a visual interface.
9. The method for monitoring the abnormal people gathering behavior in the community according to claim 8, wherein the step of obtaining the information of the residents with abnormal electricity and water consumption comprises the following steps:
the front-end water inductance sensing equipment acquires the electricity consumption and water consumption of the household every one hour, stores the data into a second database and sends the data to a second data processing and analyzing module;
the data processing and analyzing module processes and analyzes the two pairs of data and stores the result into a household abnormal electricity consumption database;
and displaying the processing result on a visual interface.
10. The method for monitoring the abnormal people gathering behavior in the community according to claim 9, wherein the step of screening out the residents with the abnormal gathering of people with the same type attribute and the abnormal household electricity and water consumption comprises the following steps:
the data processing and analyzing module is used for inquiring data in the similar attribute abnormal personnel gathering database and the household power and water consumption abnormal database in a triple-combination manner, screening out the conditions that similar attribute abnormal gathering behaviors exist in a certain building and a certain household has abnormal power and water consumption, and storing the data in the suspected abnormal gathering behavior database;
and displaying the data in the database of the abnormal gathering behavior of the suspected personnel on a visual interface.
CN202110442521.2A 2021-04-23 2021-04-23 System and method for monitoring abnormal gathering behavior of people in community Pending CN113239742A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110442521.2A CN113239742A (en) 2021-04-23 2021-04-23 System and method for monitoring abnormal gathering behavior of people in community

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110442521.2A CN113239742A (en) 2021-04-23 2021-04-23 System and method for monitoring abnormal gathering behavior of people in community

Publications (1)

Publication Number Publication Date
CN113239742A true CN113239742A (en) 2021-08-10

Family

ID=77128973

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110442521.2A Pending CN113239742A (en) 2021-04-23 2021-04-23 System and method for monitoring abnormal gathering behavior of people in community

Country Status (1)

Country Link
CN (1) CN113239742A (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110246074A (en) * 2019-05-27 2019-09-17 武汉虹信技术服务有限责任公司 A kind of community security protection management system and method based on multidimensional acquisition
CN110705451A (en) * 2019-09-27 2020-01-17 支付宝(杭州)信息技术有限公司 Face recognition method, face recognition device, terminal and server
CN112132045A (en) * 2020-09-24 2020-12-25 天津锋物科技有限公司 Community personnel abnormal behavior monitoring scheme based on computer vision
WO2021004112A1 (en) * 2019-07-05 2021-01-14 深圳壹账通智能科技有限公司 Anomalous face detection method, anomaly identification method, device, apparatus, and medium

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110246074A (en) * 2019-05-27 2019-09-17 武汉虹信技术服务有限责任公司 A kind of community security protection management system and method based on multidimensional acquisition
WO2021004112A1 (en) * 2019-07-05 2021-01-14 深圳壹账通智能科技有限公司 Anomalous face detection method, anomaly identification method, device, apparatus, and medium
CN110705451A (en) * 2019-09-27 2020-01-17 支付宝(杭州)信息技术有限公司 Face recognition method, face recognition device, terminal and server
CN112132045A (en) * 2020-09-24 2020-12-25 天津锋物科技有限公司 Community personnel abnormal behavior monitoring scheme based on computer vision

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
JOSEPH REDMON 等: "You Only Look Once：Unified, Real-Time Object Detection", 《ARXIV》 *
YUANYUAN XU 等: "CenterFace: Joint Face Detection and Alignment Using Face as Point", 《ARXIV》 *
李智慧等: "智能社区门禁系统的设计与实现", 《黑龙江大学工程学报》 *

Similar Documents

Publication Publication Date Title
CN110738127B (en) Helmet identification method based on unsupervised deep learning neural network algorithm
CN110428522B (en) Intelligent security system of wisdom new town
US9852342B2 (en) Surveillance system
CN102163290B (en) Method for modeling abnormal events in multi-visual angle video monitoring based on temporal-spatial correlation information
CN111047818A (en) Forest fire early warning system based on video image
CN106128022B (en) A kind of wisdom gold eyeball identification violent action alarm method
CN111881730A (en) Wearing detection method for on-site safety helmet of thermal power plant
US20190213424A1 (en) Image processing system and image processing method
CN110728252B (en) Face detection method applied to regional personnel motion trail monitoring
CN111582129A (en) Real-time monitoring and alarming method and device for working state of shield machine driver
CN111222478A (en) Construction site safety protection detection method and system
CN111582209A (en) Abnormal behavior supervision method for construction personnel in capital construction site
CN112163572A (en) Method and device for identifying object
CN112084963B (en) Monitoring early warning method, system and storage medium
CN112183472A (en) Method for detecting whether test field personnel wear work clothes or not based on improved RetinaNet
CN117035419B (en) Intelligent management system and method for enterprise project implementation
CN114782897A (en) Dangerous behavior detection method and system based on machine vision and deep learning
CN115691034A (en) Intelligent household abnormal condition warning method, system and storage medium
CN111767881A (en) Self-adaptive crowd density estimation device based on AI technology
CN113052139A (en) Deep learning double-flow network-based climbing behavior detection method and system
CN115309871B (en) Industrial big data processing method and system based on artificial intelligence algorithm
CN116959099A (en) Abnormal behavior identification method based on space-time diagram convolutional neural network
CN113239742A (en) System and method for monitoring abnormal gathering behavior of people in community
Peng et al. Helmet wearing recognition of construction workers using convolutional neural network
CN114283367B (en) Artificial intelligent open fire detection method and system for garden fire early warning

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20210810