CN113191196A - Novel track analysis method and system in intelligent security system - Google Patents

Novel track analysis method and system in intelligent security system Download PDF

Info

Publication number
CN113191196A
CN113191196A CN202110355459.3A CN202110355459A CN113191196A CN 113191196 A CN113191196 A CN 113191196A CN 202110355459 A CN202110355459 A CN 202110355459A CN 113191196 A CN113191196 A CN 113191196A
Authority
CN
China
Prior art keywords
track
security system
video stream
real
longitude
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
CN202110355459.3A
Other languages
Chinese (zh)
Inventor
陆倩雯
罗鑫
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.)
Beijing Ruixin High Throughput Technology Co ltd
Original Assignee
Beijing Ruixin High Throughput Technology Co ltd
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 Beijing Ruixin High Throughput Technology Co ltd filed Critical Beijing Ruixin High Throughput Technology Co ltd
Priority to CN202110355459.3A priority Critical patent/CN113191196A/en
Publication of CN113191196A publication Critical patent/CN113191196A/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
    • G06V40/168—Feature extraction; Face representation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • 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
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Computer Interaction (AREA)
  • Image Analysis (AREA)
  • Alarm Systems (AREA)

Abstract

The invention discloses a track analysis method and a track analysis system in a novel intelligent security system, wherein the method comprises the following steps: step S1: acquiring a real-time video stream; step S2: decoding the video stream; step S3: comparing and analyzing to obtain the longitude and latitude position information; step S4: and (3) sending the latitude and longitude position information into a hypothesis function obtained by repeatedly training a large number of training sets through a learning algorithm to obtain a predicted track infinitely close to a real track. The method solves the problems of poor track prediction accuracy, low reliability and weak real-time performance in the urban security system, reduces the manual intervention and calculation overhead, and improves the availability and the prediction effect of the system.

Description

Novel track analysis method and system in intelligent security system
Technical Field
The invention relates to the field of safety prevention and control, in particular to a novel track analysis method and system in an intelligent security system.
Background
The 'golden pupil' city security system is constructed for the field of safe communities, and realizes the intelligent, networking and comprehensive coverage of the intelligent community system for placing, managing and controlling the large joint work. The system has the functions of grid management, real-time early warning, tracing record, data analysis, high-definition coverage and the like.
The 'golden pupil' city security system adopts an end-tube-cloud architecture mode, can track and analyze security events such as strangers, stranger vehicles, target pedestrians, target vehicles and the like through network protocol transmission, collects data information at an edge end, transmits the data information to a high-throughput data center through a high-speed network, completes human body recognition, face recognition, vehicle recognition, defense arrangement early warning and the like, and has very important special significance for guaranteeing community safety.
Common track tracking methods in the field of security protection include a tracking method based on a gradient Histogram (HOG) feature trace, a particle filter tracking method based on a weighted color histogram model, and the like, and the methods have the problems of difficult positioning, low positioning accuracy, high calculation overhead, and the like. The requirements of high real-time performance, strong reminding, traceability and accuracy of the 'golden pupil' city security system cannot be met.
Therefore, how to solve the above technical problems is the research direction of those skilled in the art.
Disclosure of Invention
Problems to be solved by the invention
In order to solve the problems, aiming at the characteristics of the 'golden pupil' urban security system, the invention provides a novel track analysis method and system in an intelligent security system, and solves the problems of poor track prediction accuracy, low reliability and weak real-time property in the urban security system, so that the requirements of the 'golden pupil' urban security system on the real-time property and the accuracy are met.
Means for solving the problems
In order to achieve the above object, one aspect of the present invention is a trajectory analysis method in a novel intelligent security system, including the following steps:
step S1: acquiring a real-time video stream;
step S2: decoding the video stream;
step S3: comparing and analyzing to obtain the longitude and latitude position information;
step S4: and (3) sending the latitude and longitude position information into a hypothesis function obtained by repeatedly training a large number of training sets through a learning algorithm to obtain a predicted track infinitely close to a real track.
Preferably, in step S1, a monitoring device is deployed to wirelessly obtain the real-time video stream through the IP address.
Preferably, in step S2, the acquired video stream is decoded and decomposed into video picture frames.
Preferably, in step S3, the video picture frame is sent to a face recognition comparison module, and the face recognition result, the picture information, the corresponding monitoring device number, and the longitude and latitude position information are output.
Preferably, in step S3, the following sub-steps are included:
step S31: carrying out face detection on the picture;
step S32: carrying out face tracking;
step S33: and after the face features are extracted, comparing the face features with a face database, and finally outputting a comparison result.
Preferably, in step S4, the monitored longitude and latitude position information is fed into an assumption function, which is provided by a large number of real pedestrian or vehicle trajectories to a learning algorithm, and the algorithm generates an output function, which is the assumption function h.
Preferably, the longitude and latitude position information in the step S3 is sent to a feature matrix X as described below, and the value θ is continuously updated by using a gradient descent algorithm to obtain an optimal solution θ, that is, a target face trajectory or a target license plate trajectory under data analysis can be drawn
Figure BDA0003003572300000031
Preferably, in step S3, the values are continuously updated until convergence using a batch gradient descent algorithm to obtain a predicted trajectory that is infinitely close to the actual trajectory
Figure BDA0003003572300000032
Figure BDA0003003572300000033
Figure BDA0003003572300000034
In order to achieve the above object, the present invention further provides a trajectory analysis system in a novel intelligent security system, including:
a video stream acquisition module: the method comprises the steps of acquiring a real-time video stream, and wirelessly acquiring the real-time video stream through an IP address;
a video stream decoding module: it decodes the acquired video stream and decomposes it into video picture frames;
a comparison analysis module: the system is used for carrying out comparison analysis and obtaining the position information of the monitored longitude and latitude; and
a trajectory prediction module: the method is used for sending longitude and latitude position information into an assumed function obtained by repeatedly training a large number of training sets through a learning algorithm to obtain a predicted track infinitely close to a real track.
ADVANTAGEOUS EFFECTS OF INVENTION
The track analysis method and the track analysis system in the novel intelligent security system reduce the manual intervention and the calculation overhead, and improve the usability and the prediction effect of the system.
Drawings
Fig. 1 is a flowchart of a trajectory analysis method in the novel intelligent security system of the present invention.
Fig. 2 is a schematic flow chart of a trajectory analysis method in the novel intelligent security system of the present invention.
FIG. 3 is a schematic diagram of a "golden pupil" city security system.
Fig. 4 is a block diagram of a trajectory analysis system in the novel intelligent security system of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without inventive effort based on the embodiments of the present invention, are within the scope of the present invention. It should be further emphasized here that the following embodiments provide preferred embodiments, and that the various aspects (embodiments) may be used in combination or cooperation with each other.
As shown in fig. 1, which is a flowchart of a trajectory analysis method in the novel intelligent security system of the present invention, the trajectory analysis method in the novel intelligent security system of the present invention includes the following steps:
step S1: acquiring a real-time video stream;
step S2: decoding the video stream;
step S3: comparing and analyzing to obtain the longitude and latitude position information;
step S4: and (3) sending the latitude and longitude position information into a hypothesis function obtained by repeatedly training a large number of training sets through a learning algorithm to obtain a predicted track infinitely close to a real track.
The track analysis method in the novel intelligent security system is described in detail with reference to fig. 2, and the track analysis method in the novel intelligent security system specifically includes:
step S1: a real-time video stream is acquired. In order to facilitate the deployment of outdoor monitoring equipment, a host is separated from the monitoring equipment, and real-time video streams are wirelessly acquired through IP addresses.
Step S2: and (5) video decoding. And decoding the acquired video stream and decomposing the video stream into video picture frames.
Step S3: and comparing and analyzing to obtain the longitude and latitude position information. And sending the video picture frame into a face recognition comparison module, and outputting a face recognition result, picture information, a corresponding monitoring equipment number, longitude and latitude position information and the like.
Step S4: and (3) sending the latitude and longitude position information into a hypothesis function obtained by repeatedly training a large number of training sets through a learning algorithm to obtain a predicted track infinitely close to a real track.
In step S4, the monitored longitude and latitude position information is fed into the hypothesis function (generally hShown) in (a). The hypothesis function is provided by a large number of real pedestrian or vehicle tracks to a learning algorithm, and an output function generated by the algorithm is the hypothesis function h. Assuming that the input of the function h is usually denoted by x and the output by y, (x, y) denotes one sample. The hypothetical function is linearly represented: h isθ(x)＝θ0+θ1x
There may be multiple factors, x, affecting trajectory analysis1Indicating the position of the monitoring device, x2Indicating the snap definition, x, of the monitoring device3Representing monitoring device configuration parameters, etc., then the assumption can be written as:
hθ(x)＝θ0x0+θ1x1+θ2x2+...+θnxnwherein x is0＝1。
Namely, it is
Figure BDA0003003572300000061
Wherein x0＝1。
If there are multiple records, then hθ(x(i))＝θ0x0 (i)+θ1x1 (i)+θ2x2 (i)+......+θnxn (i)Corresponds to y(i)Representing the ith output.
Finally, the minimum min (J (θ)) is found, and θ is chosen to minimize the squared difference between h (x) and y. And because there are m training samples, it is necessary to calculate the square difference of each sample, and finally multiply the result by 1/2 for simplification, i.e.
Figure BDA0003003572300000062
Min (J (θ)) is calculated using a gradient descent algorithm.
Gradient descent algorithm:
Figure BDA0003003572300000063
where a is the step size of the gradient descent.
When there are multiple training samples, the batch gradient descent algorithm:
repeat until convergence (Repeat unitary conversion):
Figure BDA0003003572300000064
alpha: speed of learning, parameters given manually. Too small a setting, long convergence time, too large a setting, may exceed the minimum.
Where θ is a positive integer between 0 and 1, meaning that there are numerous components xi(where i takes 0, 1, 2.. times.n), the influence on the overall function. Such as component x1Representing the position of the surveillance camera, component x2Representing the degree of positional deviation, component xnPossibly representing manually acquired latitude and longitude, etc., which have some effect on the prediction of the trajectory. ThetaiThe closer to 1 the value of (b), the corresponding component xiThe greater the influence on the function, if thetaiThe closer to 0 the value of (b), the corresponding component xiThe less influence on the function.
Then use hθ(x)＝θ0x0+θ1x1+θ2x2+...+θn xnWherein x is01, represents the predicted trajectory of the target face or vehicle. Wherein the parameter thetaiWrite a θ column vector as follows:
Figure BDA0003003572300000071
component xiMay be a group, i.e. an X row vector (X)0，x1，x2，...，xn) When the predicted track can be written as hθ(x) X θ. The components may also be in groups, i.e. the components are in a matrix, which is as follows:
Figure BDA0003003572300000072
each component quantity is multiplied by the column vectorA predicted trajectory hθ(xi) Wherein i is 0, 1.. times.m.
Figure BDA0003003572300000073
Corresponding to each predicted track hθ(xi) All have an actual trajectory y(i)Actual trajectory y(i)And the predicted trajectory hθ(xi) With deviation, taking the sum of squares (h) for convenient calculationθ(xi)-y(i))2Where i-1 represents a group of data, i represents a plurality of groups of data from 1 to m, and a cost function is used
Figure BDA0003003572300000081
Representing the difference between the predicted trajectory and the actual trajectory, the cost function J (θ)0,θ1,...,θn) The smaller the value is, the closer the predicted track is to the actual track.
As shown in fig. 3, the "golden pupil" city security system is a block diagram, and in the "golden pupil" city security system, an access layer, a data layer, and an application layer are respectively arranged from bottom to top. The access layer comprises security video access and IOT (Internet of things) access, the data layer receives video streams provided by the equipment access layer, decodes the video streams, identifies people, vehicles, things, objects, organizations and the like, and transmits detection results to the application layer through the comparison system. The method comprises the following specific steps:
step S1: and deploying common monitoring equipment such as Haikang or Dahua, and setting to acquire real-time video stream through IP wireless after deployment. The video camera transmits data by using a code stream, different code streams are transmitted by RTSP, and the Haikang RTSP protocol address comprises an IP address, a user name, a password, a port number, a video coding mode, a channel number and a code stream type, namely a main code stream and a sub code stream.
Step S2: and the data layer of the 'golden pupil' city security system decodes the video stream in the step S1 in real time through FFMPEG.
Step S3: after the video stream is decoded into a video picture frame, the video picture frame is sent to a comparison module in a security system to identify specific face information, and information such as a corresponding monitoring equipment number, a longitude and latitude position and the like.
In this step, the following substeps are included:
step S31: carrying out face detection on the picture; the image is subjected to face detection through a face comparison program, the input of a face detection algorithm is a picture, the output is a face frame coordinate series, and the process is basically a process of scanning and judging, namely the process of scanning the face detection algorithm in an image range and judging whether a candidate area is a face one by one.
Step S32: and (4) carrying out face tracking, wherein the input is 'one face picture' and 'a face coordinate frame', and the coordinate series of the key points of the five sense organs are output. The face feature extraction is a process of converting a face image into a string of data with fixed length, and the numerical string is called as 'face feature' and has the capability of representing the face feature of the person.
Step S33: extracting the face features and then comparing the face features with a face database, wherein the input of a face comparison algorithm is two face features, the output is the similarity between the two face features, and finally, a comparison result is output.
Step S4: and (4) sending the longitude and latitude position information in the step (S3) into a feature matrix X as described below, and continuously updating the value theta by using a gradient descent algorithm to obtain an optimal solution theta, namely drawing a target face track or a target license plate track under data analysis.
Figure BDA0003003572300000091
Further, with the following batch gradient descent algorithm, the values are continuously updated until convergence, thereby obtaining a predicted trajectory infinitely close to the actual trajectory.
Repeat until convergence:
Figure BDA0003003572300000101
As shown in fig. 4, the present invention further provides a trajectory analysis system in a novel intelligent security system, where the system 1 includes:
the video stream acquisition module 11: the method comprises the steps of acquiring a real-time video stream, and wirelessly acquiring the real-time video stream through an IP address;
video stream decoding module 12: it decodes the acquired video stream and decomposes it into video picture frames;
an alignment analysis module 13: the system is used for carrying out comparison analysis and obtaining the position information of the monitored longitude and latitude; and
the trajectory prediction module 14: the method is used for sending longitude and latitude position information into an assumed function obtained by repeatedly training a large number of training sets through a learning algorithm to obtain a predicted track infinitely close to a real track.
The present invention may have various modifications within the scope not departing from the spirit of the invention, such as: the system supports detection of the mask scheme, and protocol parameters set by the system for receiving real-time data streams and the like can be changed in different implementations. Such variations are also intended to be included within the scope of the invention as claimed.
In conclusion, the track analysis method in the novel intelligent security system is a track analysis method in a 'golden pupil' urban security system, solves the problems of poor track prediction accuracy, low reliability and weak real-time performance in the urban security system, reduces manual intervention and calculation overhead, and improves system availability and prediction effect.
Those of ordinary skill in the art will understand that: the figures are merely schematic representations of one embodiment, and the blocks or flow diagrams in the figures are not necessarily required to practice the present invention.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (9)

1. A track analysis method in a novel intelligent security system is characterized by comprising the following steps:
step S1: acquiring a real-time video stream;
step S2: decoding the video stream;
step S3: comparing and analyzing to obtain the longitude and latitude position information;
step S4: and (3) sending the latitude and longitude position information into a hypothesis function obtained by repeatedly training a large number of training sets through a learning algorithm to obtain a predicted track infinitely close to a real track.
2. The novel trajectory analysis method for the intelligent security system according to claim 1, wherein in step S1, a monitoring device is deployed, and a real-time video stream is wirelessly obtained through an IP address.
3. The novel trajectory analysis method in the intelligent security system according to claim 1, wherein in step S2, the obtained video stream is decoded and decomposed into video picture frames.
4. The method for analyzing the track in the intelligent security system according to claim 1, wherein in step S3, the video picture frame is sent to the face recognition comparison module, and the face recognition result, the picture information, the corresponding monitoring device number, and the longitude and latitude position information are output.
5. The novel trajectory analysis method in the intelligent security system according to claim 1, wherein in step S3, the method comprises the following substeps:
step S31: carrying out face detection on the picture;
step S32: carrying out face tracking;
step S33: and after the face features are extracted, comparing the face features with a face database, and finally outputting a comparison result.
6. The trajectory analysis method in the novel intelligent security system according to claim 1, wherein in step S4, the longitude and latitude position information is monitored and sent to an assumption function, the assumption function is provided to a learning algorithm by a large number of real pedestrian or vehicle trajectories, and an output function generated by the algorithm is an assumption function h.
7. The novel track analysis method in the intelligent security system as claimed in claim 1, wherein the longitude and latitude position information in step S3 is sent to a feature matrix X as follows, and a gradient descent algorithm is used to continuously update the value θ to obtain an optimal solution θ, i.e. a target face track or a target license plate track under data analysis can be drawn
Figure FDA0003003572290000021
8. The method for analyzing the track in the intelligent security system according to claim 7, wherein in step S3, the following batch gradient descent algorithm is used, and the updated value is continuously updated until convergence, so as to obtain the predicted track infinitely close to the actual track
Figure FDA0003003572290000022
9. The utility model provides a track analytic system among novel intelligent security system which characterized in that includes:
a video stream acquisition module: the method comprises the steps of acquiring a real-time video stream, and wirelessly acquiring the real-time video stream through an IP address;
a video stream decoding module: it decodes the acquired video stream and decomposes it into video picture frames;
a comparison analysis module: the system is used for carrying out comparison analysis and obtaining the position information of the monitored longitude and latitude; and
a trajectory prediction module: the method is used for sending longitude and latitude position information into an assumed function obtained by repeatedly training a large number of training sets through a learning algorithm to obtain a predicted track infinitely close to a real track.
CN202110355459.3A 2021-04-01 2021-04-01 Novel track analysis method and system in intelligent security system Pending CN113191196A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110355459.3A CN113191196A (en) 2021-04-01 2021-04-01 Novel track analysis method and system in intelligent security system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110355459.3A CN113191196A (en) 2021-04-01 2021-04-01 Novel track analysis method and system in intelligent security system

Publications (1)

Publication Number Publication Date
CN113191196A true CN113191196A (en) 2021-07-30

Family

ID=76974709

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110355459.3A Pending CN113191196A (en) 2021-04-01 2021-04-01 Novel track analysis method and system in intelligent security system

Country Status (1)

Country Link
CN (1) CN113191196A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114257790A (en) * 2021-12-21 2022-03-29 北京睿芯高通量科技有限公司 Mask early warning method and device in intelligent security system, electronic equipment and storage medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109326124A (en) * 2018-10-17 2019-02-12 江西洪都航空工业集团有限责任公司 A kind of urban environment based on machine vision parks cars Activity recognition system
CN109684916A (en) * 2018-11-13 2019-04-26 恒睿(重庆)人工智能技术研究院有限公司 Based on path locus data exception detection method, system, equipment and storage medium
WO2019184125A1 (en) * 2018-03-30 2019-10-03 平安科技（深圳）有限公司 Micro-expression-based risk identification method and device, equipment and medium
CN111008574A (en) * 2019-11-22 2020-04-14 智慧视通(杭州)科技发展有限公司 Key person track analysis method based on body shape recognition technology

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019184125A1 (en) * 2018-03-30 2019-10-03 平安科技（深圳）有限公司 Micro-expression-based risk identification method and device, equipment and medium
CN109326124A (en) * 2018-10-17 2019-02-12 江西洪都航空工业集团有限责任公司 A kind of urban environment based on machine vision parks cars Activity recognition system
CN109684916A (en) * 2018-11-13 2019-04-26 恒睿(重庆)人工智能技术研究院有限公司 Based on path locus data exception detection method, system, equipment and storage medium
CN111008574A (en) * 2019-11-22 2020-04-14 智慧视通(杭州)科技发展有限公司 Key person track analysis method based on body shape recognition technology

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114257790A (en) * 2021-12-21 2022-03-29 北京睿芯高通量科技有限公司 Mask early warning method and device in intelligent security system, electronic equipment and storage medium

Similar Documents

Publication Publication Date Title
CN102799935B (en) Human flow counting method based on video analysis technology
CN112396658B (en) Indoor personnel positioning method and system based on video
CN102610102B (en) Suspect vehicle inspection and control method and system
CN109657575B (en) Intelligent video tracking algorithm for outdoor constructors
CN101799876B (en) Video/audio intelligent analysis management control system
CN113723257A (en) Event short video generation method, system, device and storage medium
CN107197199A (en) A kind of intelligent monitoring and controlling device and method for tracking target
CN111008574A (en) Key person track analysis method based on body shape recognition technology
Jiang et al. An Approach for Crowd Density and Crowd Size Estimation.
CN105657372A (en) Method and system for realizing intelligent detection and early warning of on-duty guard posts by videos
CN116778673A (en) Water area safety monitoring method, system, terminal and storage medium
CN113191196A (en) Novel track analysis method and system in intelligent security system
CN113191273A (en) Oil field well site video target detection and identification method and system based on neural network
CN108965804A (en) A kind of video structural technology for city security protection
CN113569656B (en) Examination room monitoring method based on deep learning
CN108460357A (en) A kind of windowing alarm detection system and method based on image recognition
CN113378633A (en) Method and system for detecting quality of streaming media signal
CN112738471A (en) Security monitoring data processing method based on Internet of things and artificial intelligence and cloud communication server
CN110533889B (en) Sensitive area electronic equipment monitoring and positioning device and method
CN111800428A (en) Real-time statistical method and system for digital conference participation
CN109344705B (en) Pedestrian behavior detection method and system
CN116152745A (en) Smoking behavior detection method, device, equipment and storage medium
Xu et al. Unsupervised abnormal behaviour detection with overhead crowd video
CN206948499U (en) The monitoring of student's real training video frequency tracking, evaluation system
CN114898250A (en) Suckling piglet multi-target tracking method based on self-attention mechanism

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

Application publication date: 20210730

RJ01 Rejection of invention patent application after publication