CN107679471A - Indoor occupant sky hilllock detection method based on video monitoring platform - Google Patents
Indoor occupant sky hilllock detection method based on video monitoring platform Download PDFInfo
- Publication number
- CN107679471A CN107679471A CN201710871443.1A CN201710871443A CN107679471A CN 107679471 A CN107679471 A CN 107679471A CN 201710871443 A CN201710871443 A CN 201710871443A CN 107679471 A CN107679471 A CN 107679471A
- Authority
- CN
- China
- Prior art keywords
- hilllock
- frame
- state
- duty
- empty
- 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
Links
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20224—Image subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30232—Surveillance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30242—Counting objects in image
Abstract
The present invention relates to a kind of Interior Space hilllock detection method based on video monitoring platform, including:To pretreated frame of video, handle to obtain the difference image of every two frame using frame differential method;A state machine is defined, the dynamic translation of four kinds of states is realized, specifically includes four kinds of states:Nobody on duty, doubtful empty hilllock on duty, doubtful, someone are on duty;The state of operator on duty is judged according to difference image average, if the difference image average is not more than 200, into state machine;Conversely, it is constant to be maintained the original state if the difference image average is more than 200, do not enter state machine, the interference brought because imaging jump in brightness in monitor video is eliminated with appropriate;Variable, the personnel's status information on duty for converting to obtain by state machine correspondingly change state variable and empty hilllock timer.
Description
Technical field
The invention belongs to field of intelligent video surveillance, specifically belongs to a kind of interior based on existing video monitoring platform
Personnel's sky hilllock detecting system and method.
Background technology
Destabilizing factor is increasing in the society of current high speed development, all trades and professions, and safety problem is got over by people
Come more concerns, especially safety-security area.In order to ensure national important safety department, the safety of important unit facility, safeguard
The long-term stability of society, post is on duty just to become an inevitable choice.If operator on duty leaves the post without authorization in watch time, very
It is likely to cause immeasurable loss.In order to prevent this situation, many enterprises and department take the mode of supervision
Empty hilllock detection is carried out, to take counter-measure in time, is eliminated safe hidden trouble.
With the continuous development of Video Supervision Technique, gradually instead of using the means of video monitoring to monitor duty information
The mode manually inspected the sentries.Traditional video monitoring system is to need special messenger to night shift room monitor video real time inspection, monitoring period
It is long, spiritual need high concentration, form higher human cost.Moreover, people watches monitoring display device attentively when reaching certain time
Scatterbrained situation just occurs, easily misses the video segment of key, is unable to reach supervision purpose.Intelligent video is supervised
Control technology is arisen at the historic moment.Relative to traditional monitoring means, intelligent video monitoring then reliability it is higher, reaction faster, cost more
It is low, by camera acquisition video image, the information characteristics of intellectual analysis video frame images, judged simultaneously according to predetermined criterion
Reaction, then it is to be determined whether according to predetermined criterion there occurs empty hilllock behavior and alarmed and taken for being detected for empty hilllock
Corresponding measure.
Empty hilllock detection based on monitor supervision platform, mainly including camera video acquisition, system background intellectual analysis and display
Device end real-time display, wherein, intellectual analysis part is related to target detection, tracking and analysis identification etc..At present, in video image
Object detection method mainly has:Frame differential method, background subtraction and optical flow method.Inter-frame difference method is most simple, most direct, greatly
It is used for that background is simple, the small situation of environmental disturbances;Background subtraction is more sensitive to the dynamic change of scene;Optical flow method can be with
Target, but unsuitable processing in real time are detected in camera motion.Generally speaking, the empty hilllock inspection based on video monitoring platform
Survey, be the video sequence obtained by real-time intellectual analysis monitoring camera, obtain the status information of operator on duty, work as appearance
Make a response, eliminate safe hidden trouble immediately during empty hilllock situation.
The existing method for solving empty hilllock test problems, it is generally based on the empty hilllock detection method of video monitoring platform.Such as
In patent CN104021653A, pass through video dynamic analysis video camera, video analysis alarm and centered video monitoring server
Early warning, alarm, SMS alarm are run, realizes the function of road junction state video analysis alarm on duty, it is mainly according to frame of video figure
Seem that no generation dynamic change judges whether that someone is on duty;Graphical analysis and track algorithm are used in patent CN102740059A
Supervise operator on duty, using recognizer identify in image whether someone, accumulating will when nobody reaches preset time parameter the time
Unwatched supervisory signals are sent to control centre, without the detailed process for clearly referring to human testing algorithm, and only
There are nobody and someone's two states.Generally speaking, if it is possible to design a kind of higher human testing algorithm and one of accuracy
The more perfect state transformation flow of set then can preferably realize the Interior Space hilllock detection function based on monitor supervision platform.
The content of the invention
It is an object of the invention to provide one kind to be based on existing video monitoring platform, can be real-time to the behavior of Interior Space hilllock
Monitor and the method for timely processing, technical scheme are as follows:
A kind of Interior Space hilllock detection method based on video monitoring platform, including following step:
1) empty hilllock detection parameters are set, include two parameters of change of scale parameter and empty hilllock time threshold;
2) input video, the video is read frame by frame, all frame of video are multiplied by into change of scale parameter is scaled to unified size,
Frame of video is pre-processed, so that subsequent step is further handled;
3) to pretreated frame of video, handle to obtain the difference image of every two frame using frame differential method, and carry out threshold
It is worth binary conversion treatment, Morphological scale-space, calculates the pixel intensity average and standard deviation of difference image, then circulation performs this mistake
Journey, dynamic prospect is detected in real time;
4) state machine is defined, the dynamic translation of four kinds of states is realized, specifically includes four kinds of states:Nobody is on duty, doubtful
It is on duty like on duty, doubtful empty hilllock, someone;
5) state of operator on duty is judged according to the above-mentioned difference image average being calculated, if the difference image average
No more than 200, then into state machine;Conversely, it is constant to be maintained the original state if the difference image average is more than 200, do not enter shape
State machine, the interference brought because imaging jump in brightness in monitor video is eliminated with appropriate;
6) state machine parameter initialization is realized, defines an empty hilllock timer for being initialized as 0 and a state variable, is led to
Cross personnel's status information on duty that state machine converts to obtain and correspondingly change state variable and empty hilllock timer;
7) running status machine, every 10 frame judge a state change, to realize the dynamic translation of four kinds of states;In wherein
During any one state, first have to be loaded into the human testing grader operation human testing process of pre-training, then correspond to and obtain
Personnel's Information Statistics on duty in every 10 frame video image, the current state of personnel is judged according to corresponding decision criteria and divided
Analyse state transformation trend;Wherein,
Human testing grader is trained as training sample using upper half of human body sample, human testing process is as follows:
Current pretreated frame of video is detected using the good human testing grader of pre-training, multiple dimensioned scanning frame of video
Image, obtain being detected as the target of human body, be stored in after it is identified with rectangle frame in rectangle frame queue, then in former video figure
As upper drafting human body bounding rectangles frame.
The conversion process of four kinds of states is as follows:
During " nobody is on duty " state, empty hilllock timer starts timing, every time add up 1, run human testing process, if
Frame of video detects people, then enters the state of " doubtful on duty ", otherwise state is constant;
During " doubtful on duty " state, human testing process is run, if it is difference image average there are at least 6 frames in every 10 frame
More than 2, then it is assumed that it is kept in motion, it is on the contrary then be static;If being kept in motion, it is determined as dynamic frame, continues to count
The frame of video frame number of people is detected, if at least 6 frames detect people in 10 frames, then it is assumed that state is " someone is on duty ";Conversely,
Think " nobody is on duty ", empty hilllock timer is added 10;If in static state, that is, it is determined as static frames, now calculates every in 10 frames
The overlapping area of the rectangle frame of people between any two is detected in two frames, if overlapping area and the ratio of single rectangle frame area are not
The same people for being substantially at static sitting posture is considered less than 0.2, if at least 6 frames are in static state to be such a in 10 frames
The people of sitting posture, then it is assumed that " someone is on duty ", be otherwise " nobody is on duty ";
During " someone is on duty " state, empty hilllock timer restarts to count from 0, runs human testing process, if detection
Less than people, into the state on " doubtful empty hilllock ", otherwise state is constant;
During " doubtful empty hilllock " state, empty hilllock timer initial value is still 0, runs human testing process, and every 10 frame judges once,
If at least 8 frames are not more than 2 situation, as static frames, also, at least 6 frames are not for difference image average in every 10 frame
Detect human body, then it is assumed that state is " nobody is on duty ", and empty hilllock timer adds 10, on the contrary, it is believed that " someone is on duty ";
By that analogy, once above-mentioned state machine process, completion once judge for every 10 frame operation;
8) state machine conversion process operation finishes, and checks sky hilllock timer, is preset if the empty hilllock timer cumulative time exceedes
Empty hilllock time threshold, for example nobody is on duty in 5min, then judges, there occurs empty hilllock behavior, to carry out Realtime Alerts and display alarm shape
State information, empty hilllock timer is reset to 0, restarts timing;Conversely, it is judged to that empty hilllock behavior does not occur;
9) each testing staff state on duty is last, all that the queue for storing human body bounding rectangles frame in single-frame images is clear
Sky, detected for the empty hilllock of next round.
Color information of the present invention independent of video frame images, the monitor video at night can also be handled, and pass through foundation
One complete indoor occupant sky hilllock detecting system based on video monitoring platform, using accurate human testing algorithm and
Fairly perfect state machine conversion process, real-time analysis monitoring video sequence, obtains the status information of operator on duty, works as appearance
Made a response immediately during empty hilllock situation and take corresponding measure, eliminate potential safety hazard.In this way, phase is not only increased
The supervision level of pass department can also aid in related personnel preferably to complete task, reduce human cost, ensured life
Living safety is produced, it is significant for intelligent security guard field.
Brief description of the drawings
The video monitoring system block diagram that Fig. 1 is carried by the inventive method
Fig. 2 is the detecting system of human body block diagram designed by the inventive method
Fig. 3 is that the state machine designed by the inventive method converts block diagram
Fig. 4 is the flow chart of the inventive method
Embodiment
The general processing framework of video monitoring system is in existing safety-security area:Passed through by the picture of analog video camera shooting
A cable part is directly transmitted to monitor and shown, another part is transmitted to DVR.Into the analog signal of DVR
It is changed into digital code stream, is on the one hand encoded, is stored in document form in DVR;On the other hand, net can be passed through
Network connects DVR at any time, and extraction code stream is shown, analyzed.As shown in Figure 1.
The Interior Space hilllock detecting system software based on video monitoring platform formed based on method proposed by the present invention, can
By network connection HD recording, to gather video data and be analyzed in real time, or HD recording is stored in by extraction
Video file in machine after transcoding.Step is as follows:Empty hilllock parameter, including change of scale parameter and empty hilllock time are set first
Threshold parameter;Video is read frame by frame and video is pre-processed;Video is handled by frame differential method to obtain difference image;
Definition status machine, four kinds of states are included in state machine:" nobody is on duty ", " doubtful on duty ", " doubtful empty hilllock ", " someone is on duty ";
By analyzing the average size of difference image, into different states;At each state, first pretreated frame of video is transported
Row human testing process, using whether there is human object in the human testing detection of classifier of the pre-training frame of video;By
State machine, which converts, realizes the mutual conversions of four kinds of states, can accurately judge " nobody is on duty ", " doubtful on duty ", " doubt
Like empty hilllock ", " someone is on duty " this several state, the corresponding value for counting empty hilllock timer;If empty hilllock time timer timing surpasses
Cross default empty hilllock time threshold then to be alarmed, timing will be restarted again after its zero setting;By that analogy, every 10 frame judges one
It is secondary;Result video after most handling at last preserves.
By real-time monitor video image, become using the higher human testing algorithm of accuracy and more perfect state
Differentiation mechanism is changed, detection in real time and the state on duty of analysis personnel, effective supervision to the behavior of Interior Space hilllock is realized, eliminates
Potential safety hazard.It is of the invention specific as shown in Figure 4.
Various pieces are described in detail below:
1. empty hilllock parameter setting
Change of scale parameter, such as 0.75 are set, for post-processing;Another parameter is sky hilllock time threshold, than
Such as 5min, so that alarm decision is detected on later stage sky hilllock.
2. frame of video pre-processes
The video of input is read frame by frame, and change of scale parameter, such as 0.75 are multiplied by each frame of video, dwindles into original chi
Very little 0.75 times, computational efficiency can be improved, and the frame of video after scaled is pre-processed:Because some monitor videos are
Night video, so frame of video is converted into gray scale bitmap-format by rgb signal, so can be with independent of color information.
3. dynamic foreground extraction
To pretreated frame of video, handle to obtain the difference image of every two frame using frame differential method, and carry out threshold value
Binary conversion treatment, Morphological scale-space, the pixel intensity average and standard deviation of difference image being calculated, then circulation performs this process,
Detection dynamic prospect in real time.
4. definition status machine
A state machine is defined, the dynamic translation of four kinds of states is realized, specifically includes four kinds of states:Nobody is on duty, doubtful
On duty, doubtful empty hilllock, someone are on duty.
5. enter state machine
The state of operator on duty is judged according to the above-mentioned difference image average being calculated, if the difference image average is not
More than 200, then into state machine;Conversely, it is constant to be maintained the original state if the difference image average is more than 200, do not enter state
Machine, it can so eliminate the interference brought because imaging jump in brightness in monitor video.
6. state machine parameter initialization
Define the empty hilllock timer and a state variable that an initial value is 0, the personnel's value for converting to obtain by state machine
Class's status information correspondingly adjusts the state variable and the value of empty hilllock timer.
7. state machine is changed
Running status machine, testing staff's state on duty, every 10 frame judge a state change, realize the dynamic of four kinds of states
Conversion;During in any of which state, first have to be loaded into the human testing grader operation human testing process of pre-training,
Then correspond to and obtain personnel's Information Statistics on duty in every 10 frame video image, working as personnel is judged according to corresponding decision criteria
Preceding state and analysis state transformation trend;It is specifically described below:
Wherein, human testing process is as shown in Fig. 2 comprise the following steps:
In view of the specific environment of personnel's night shift room detection, most of is the people of sitting posture, and occasionally there are the feelings that personnel walk about
Condition, but can substantially ensure the integrality of upper half of human body, and upper half of human body contains head-and-shoulder area and correspondingly
The higher architectural feature of human bioequivalence rate, based on such consideration, upper half of human body sample used herein is as training sample
Original training human testing grader.Main human testing step is as follows:Using the human testing grader of pre-training to working as
Preceding pretreated frame of video is detected, multiple dimensioned scanning video frame images, until whole video frame images of traversal, are obtained
The target of human body is detected as, is stored in after it is identified with rectangle frame in rectangle frame queue, is then drawn on original video image
(rectangle frame parameter corresponds to bounding rectangles frame by amplification divided by predefined change of scale parameter carries out scaling, so as to adapt to
Artwork size);
In addition, state machine FB(flow block) as shown in figure 3, four kinds of states conversion process approximately as:
During " nobody is on duty " state, empty hilllock timer starts timing, every time add up 1, run human testing process, if
Frame of video detects people, then enters the state of " doubtful on duty ", otherwise state is constant;
During " doubtful on duty " state, human testing process is run, if it is difference image average there are at least 6 frames in every 10 frame
More than 2, then it is assumed that be kept in motion, otherwise be static state;If being kept in motion, it is determined as dynamic frame, continues statistics inspection
The frame of video frame number of people is measured, if at least 6 frames detect people in 10 frames, then it is assumed that state is " someone is on duty ";Conversely, recognize
For " nobody is on duty ", empty hilllock timer adds 10;If in static state, that is, it is determined as static frames, now calculates in 10 frames in every two frame
The overlapping area of the rectangle frame of people between any two is detected, if overlapping area and the ratio of single rectangle frame area are not less than
0.2 is considered the same people for being substantially at static sitting posture, if at least 6 frames are in static sitting posture to be such a in 10 frames
People, then it is assumed that " someone is on duty ", be otherwise " nobody is on duty ";
During " someone is on duty " state, empty hilllock timer counts from 0 again, runs human testing process, if detection suddenly
Less than people, into the state on " doubtful empty hilllock ", otherwise state is constant;
During " doubtful empty hilllock " state, empty hilllock timer initial value is still 0, runs human testing process, and every 10 frame judges once,
If at least 8 frames are not more than 2 situation, as static frames, also, at least 6 frames are not for difference image average in every 10 frame
Detect human body, then it is assumed that state is " nobody is on duty ", and empty hilllock timer adds 10, on the contrary, it is believed that " someone is on duty ";
By that analogy, once above-mentioned state machine process, completion once judge for every 10 frame operation;
8. empty hilllock behavior judges
The whole conversion process operation of state machine finishes, and sky hilllock timer is checked, if the empty hilllock timer cumulative time exceedes
Empty hilllock time threshold is preset, for example nobody is on duty in 5min, then judges there occurs empty hilllock behavior, is alarmed in real time and shows report
Alert status information, resets to 0 by empty hilllock timer, restarts timing;Conversely, it is judged to that empty hilllock behavior does not occur.
9. circulation carries out empty hilllock detection
Testing staff's state on duty is last every time, all that the queue for storing human body bounding rectangles frame in single-frame images is clear
Sky, detected for the empty hilllock of next round.
10. preserve result video
After whole system engineering operation, whole video processing procedures can be stored in the form of new video, with
Checked in the future for staff.
Claims (1)
1. a kind of Interior Space hilllock detection method based on video monitoring platform, including following step:
1) empty hilllock detection parameters are set, include two parameters of change of scale parameter and empty hilllock time threshold;
2) input video, read the video frame by frame, all frame of video are multiplied by into change of scale parameter is scaled to unified size, to regarding
Frequency frame is pre-processed, so that subsequent step is further handled;
3) to pretreated frame of video, handle to obtain the difference image of every two frame using frame differential method, and carry out threshold value two
Value processing, Morphological scale-space, the pixel intensity average and standard deviation of difference image are calculated, then circulation performs this process, real
When detect dynamic prospect;
4) state machine is defined, the dynamic translation of four kinds of states is realized, specifically includes four kinds of states:Nobody it is on duty, doubtful
Hilllock, doubtful empty hilllock, someone are on duty;
5) state of operator on duty is judged according to the above-mentioned difference image average being calculated, if the difference image average is little
In 200, then into state machine;Conversely, it is constant to be maintained the original state if the difference image average is more than 200, do not enter state machine,
The interference brought because imaging jump in brightness in monitor video is eliminated with appropriate;
6) state machine parameter initialization is realized, an empty hilllock timer for being initialized as 0 and a state variable is defined, passes through shape
Personnel's status information on duty that state machine converts to obtain correspondingly changes state variable and empty hilllock timer;
7) running status machine, every 10 frame judge a state change, to realize the dynamic translation of four kinds of states;In wherein any
During a kind of state, first have to be loaded into the human testing grader operation human testing process of pre-training, then correspond to and obtain every 10
Personnel's Information Statistics on duty in frame video image, the current state and analysis shape of personnel are judged according to corresponding decision criteria
State converts trend;Wherein,
Human testing grader is trained as training sample using upper half of human body sample, human testing process is as follows:Use
The good human testing grader of pre-training detects to current pretreated frame of video, multiple dimensioned scanning frame of video figure
Picture, obtain being detected as the target of human body, be stored in after it is identified with rectangle frame in rectangle frame queue, then in original video image
Upper drafting human body bounding rectangles frame.
The conversion process of four kinds of states is as follows:
During " nobody is on duty " state, empty hilllock timer starts timing, every time add up 1, run human testing process, if video
Frame detects people, then enters the state of " doubtful on duty ", otherwise state is constant;
During " doubtful on duty " state, human testing process is run, if there are at least 6 frames to be more than for difference image average in every 10 frame
2, then it is assumed that it is kept in motion, it is on the contrary then be static;If being kept in motion, it is determined as dynamic frame, continues statistic mixed-state
To the frame of video frame number of people, if at least 6 frames detect people in 10 frames, then it is assumed that state is " someone is on duty ";It is on the contrary, it is believed that
" nobody is on duty ", empty hilllock timer is added 10;If in static state, that is, it is determined as static frames, now calculates every two frame in 10 frames
In detect the overlapping area of the rectangle frame of people between any two, if overlapping area and the ratio of single rectangle frame area are not less than
0.2 is considered the same people for being substantially at static sitting posture, if at least 6 frames are in static sitting posture to be such a in 10 frames
People, then it is assumed that " someone is on duty ", be otherwise " nobody is on duty ";
During " someone is on duty " state, empty hilllock timer restarts to count from 0, human testing process is run, if can't detect
People, into the state on " doubtful empty hilllock ", otherwise state is constant;
During " doubtful empty hilllock " state, empty hilllock timer initial value is still 0, runs human testing process, and every 10 frame judges once, if
At least 8 frames are not more than 2 situation, as static frames, also, at least 6 frames do not detect for difference image average in every 10 frame
To human body, then it is assumed that state is " nobody is on duty ", and empty hilllock timer adds 10, on the contrary, it is believed that " someone is on duty ";
By that analogy, once above-mentioned state machine process, completion once judge for every 10 frame operation;
8) state machine conversion process operation finishes, and checks sky hilllock timer, if the empty hilllock timer cumulative time, which exceedes, presets empty hilllock
Nobody is on duty in time threshold, such as 5min, then judges there occurs empty hilllock behavior, carries out Realtime Alerts and display alarm state letter
Breath, resets to 0 by empty hilllock timer, restarts timing;Conversely, it is judged to that empty hilllock behavior does not occur;
9) each testing staff state on duty is last, all empties the queue for storing human body bounding rectangles frame in single-frame images,
Detected for the empty hilllock of next round.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710871443.1A CN107679471B (en) | 2017-09-24 | 2017-09-24 | Indoor personnel air post detection method based on video monitoring platform |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710871443.1A CN107679471B (en) | 2017-09-24 | 2017-09-24 | Indoor personnel air post detection method based on video monitoring platform |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107679471A true CN107679471A (en) | 2018-02-09 |
CN107679471B CN107679471B (en) | 2020-03-06 |
Family
ID=61138000
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710871443.1A Expired - Fee Related CN107679471B (en) | 2017-09-24 | 2017-09-24 | Indoor personnel air post detection method based on video monitoring platform |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107679471B (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108564050A (en) * | 2018-04-23 | 2018-09-21 | 西安安邦鼎立智能科技有限公司 | A kind of two visitors one based on the video enterprise supervision personnel that endanger inspect the sentries method and system |
CN109190710A (en) * | 2018-09-13 | 2019-01-11 | 东北大学 | Detection method of leaving post based on Haar-NMF feature and cascade Adaboost classifier |
CN109492620A (en) * | 2018-12-18 | 2019-03-19 | 广东中安金狮科创有限公司 | Monitoring device and its control device, post monitoring method and readable storage medium storing program for executing |
CN109867186A (en) * | 2019-03-18 | 2019-06-11 | 浙江新再灵科技股份有限公司 | A kind of elevator malfunction detection method and system based on intelligent video analysis technology |
CN110443179A (en) * | 2019-07-29 | 2019-11-12 | 思百达物联网科技(北京)有限公司 | It leaves the post detection method, device and storage medium |
CN110633680A (en) * | 2019-09-19 | 2019-12-31 | 天津天地伟业机器人技术有限公司 | Video-based people number abnormity detection method |
CN110708823A (en) * | 2019-10-29 | 2020-01-17 | 佛山科学技术学院 | Classroom automatic power-off control method and system based on image recognition |
CN111970495A (en) * | 2020-08-21 | 2020-11-20 | 湖南工学院 | Remote automatic light control system |
CN112053224A (en) * | 2020-09-02 | 2020-12-08 | 中国银行股份有限公司 | Business processing monitoring implementation method, device and system |
CN113065386A (en) * | 2020-12-25 | 2021-07-02 | 泰州可以信息科技有限公司 | Preset duration worker on-duty state detection system |
CN113489929A (en) * | 2021-08-06 | 2021-10-08 | 远见智诚市场调研咨询(广东)有限公司 | Method and device for monitoring supervision of garbage classification, computer equipment and storage medium |
CN117412180A (en) * | 2023-12-15 | 2024-01-16 | 杭州三信网络技术有限公司 | Welding machine based on multi-camera linkage target monitoring and target monitoring method |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070160289A1 (en) * | 2001-03-23 | 2007-07-12 | Objectvideo, Inc. | Video segmentation using statistical pixel modeling |
CN101883257A (en) * | 2009-05-08 | 2010-11-10 | 上海弘视通信技术有限公司 | Guard detection system and detection method thereof |
CN102647580A (en) * | 2012-04-27 | 2012-08-22 | 浙江晨鹰科技有限公司 | Video monitoring method and system |
CN104408406A (en) * | 2014-11-03 | 2015-03-11 | 安徽中科大国祯信息科技有限责任公司 | Staff off-post detection method based on frame difference method and background subtraction method |
CN106570467A (en) * | 2016-10-25 | 2017-04-19 | 南京南瑞集团公司 | Convolutional neutral network-based worker absence-from-post detection method |
CN107103300A (en) * | 2017-04-22 | 2017-08-29 | 高新兴科技集团股份有限公司 | One kind is left the post detection method and system |
-
2017
- 2017-09-24 CN CN201710871443.1A patent/CN107679471B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070160289A1 (en) * | 2001-03-23 | 2007-07-12 | Objectvideo, Inc. | Video segmentation using statistical pixel modeling |
CN101883257A (en) * | 2009-05-08 | 2010-11-10 | 上海弘视通信技术有限公司 | Guard detection system and detection method thereof |
CN102647580A (en) * | 2012-04-27 | 2012-08-22 | 浙江晨鹰科技有限公司 | Video monitoring method and system |
CN104408406A (en) * | 2014-11-03 | 2015-03-11 | 安徽中科大国祯信息科技有限责任公司 | Staff off-post detection method based on frame difference method and background subtraction method |
CN106570467A (en) * | 2016-10-25 | 2017-04-19 | 南京南瑞集团公司 | Convolutional neutral network-based worker absence-from-post detection method |
CN107103300A (en) * | 2017-04-22 | 2017-08-29 | 高新兴科技集团股份有限公司 | One kind is left the post detection method and system |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108564050A (en) * | 2018-04-23 | 2018-09-21 | 西安安邦鼎立智能科技有限公司 | A kind of two visitors one based on the video enterprise supervision personnel that endanger inspect the sentries method and system |
CN109190710A (en) * | 2018-09-13 | 2019-01-11 | 东北大学 | Detection method of leaving post based on Haar-NMF feature and cascade Adaboost classifier |
CN109190710B (en) * | 2018-09-13 | 2022-04-08 | 东北大学 | off-Shift detection method based on Haar-NMF characteristics and cascade Adaboost classifier |
CN109492620A (en) * | 2018-12-18 | 2019-03-19 | 广东中安金狮科创有限公司 | Monitoring device and its control device, post monitoring method and readable storage medium storing program for executing |
CN109867186B (en) * | 2019-03-18 | 2020-11-10 | 浙江新再灵科技股份有限公司 | Elevator trapping detection method and system based on intelligent video analysis technology |
CN109867186A (en) * | 2019-03-18 | 2019-06-11 | 浙江新再灵科技股份有限公司 | A kind of elevator malfunction detection method and system based on intelligent video analysis technology |
CN110443179B (en) * | 2019-07-29 | 2021-11-12 | 思百达物联网科技(北京)有限公司 | Off-post detection method and device and storage medium |
CN110443179A (en) * | 2019-07-29 | 2019-11-12 | 思百达物联网科技(北京)有限公司 | It leaves the post detection method, device and storage medium |
CN110633680A (en) * | 2019-09-19 | 2019-12-31 | 天津天地伟业机器人技术有限公司 | Video-based people number abnormity detection method |
CN110708823A (en) * | 2019-10-29 | 2020-01-17 | 佛山科学技术学院 | Classroom automatic power-off control method and system based on image recognition |
CN111970495A (en) * | 2020-08-21 | 2020-11-20 | 湖南工学院 | Remote automatic light control system |
CN111970495B (en) * | 2020-08-21 | 2021-12-21 | 湖南工学院 | Remote automatic light control system |
CN112053224A (en) * | 2020-09-02 | 2020-12-08 | 中国银行股份有限公司 | Business processing monitoring implementation method, device and system |
CN112053224B (en) * | 2020-09-02 | 2023-08-18 | 中国银行股份有限公司 | Service processing monitoring realization method, device and system |
CN113065386A (en) * | 2020-12-25 | 2021-07-02 | 泰州可以信息科技有限公司 | Preset duration worker on-duty state detection system |
CN113489929A (en) * | 2021-08-06 | 2021-10-08 | 远见智诚市场调研咨询(广东)有限公司 | Method and device for monitoring supervision of garbage classification, computer equipment and storage medium |
CN117412180A (en) * | 2023-12-15 | 2024-01-16 | 杭州三信网络技术有限公司 | Welding machine based on multi-camera linkage target monitoring and target monitoring method |
CN117412180B (en) * | 2023-12-15 | 2024-03-15 | 杭州三信网络技术有限公司 | Welding machine based on multi-camera linkage target monitoring and target monitoring method |
Also Published As
Publication number | Publication date |
---|---|
CN107679471B (en) | 2020-03-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107679471A (en) | Indoor occupant sky hilllock detection method based on video monitoring platform | |
KR101942808B1 (en) | Apparatus for CCTV Video Analytics Based on Object-Image Recognition DCNN | |
JP4803376B2 (en) | Camera tampering detection method | |
US20060170769A1 (en) | Human and object recognition in digital video | |
CN104966304B (en) | Multi-target detection tracking based on Kalman filtering and nonparametric background model | |
CN105894702A (en) | Invasion detecting alarming system based on multi-camera data combination and detecting method thereof | |
KR20090086898A (en) | Detection of smoke with a video camera | |
AU2009210794A1 (en) | Video sensor and alarm system and method with object and event classification | |
CN106127292B (en) | Flow method of counting and equipment | |
CN109867186B (en) | Elevator trapping detection method and system based on intelligent video analysis technology | |
CN101299275A (en) | Method and device for detecting target as well as monitoring system | |
JPS6286990A (en) | Abnormality supervisory equipment | |
EP2000952B1 (en) | Smoke detecting method and device | |
CN109409289A (en) | A kind of electric operating safety supervision robot security job identifying method and system | |
CN111223263A (en) | Full-automatic comprehensive fire early warning response system | |
CN113903058A (en) | Intelligent control system based on regional personnel identification | |
CN107688793A (en) | A kind of outside transformer substation fire automatic monitoring method for early warning | |
CN111476964A (en) | Remote forest fire prevention monitoring system and method | |
CN112163572A (en) | Method and device for identifying object | |
CN116416577A (en) | Abnormality identification method for construction monitoring system | |
CN115691034A (en) | Intelligent household abnormal condition warning method, system and storage medium | |
CN113435278A (en) | Crane safety detection method and system based on YOLO | |
CN103152558A (en) | Intrusion detection method based on scene recognition | |
KR100887942B1 (en) | System for sensing abnormal phenomenon on realtime and method for controlling the same | |
KR20160093253A (en) | Video based abnormal flow detection method and system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200306 Termination date: 20200924 |
|
CF01 | Termination of patent right due to non-payment of annual fee |