CN106204659B - Elevator switch door detection method based on light stream - Google Patents
Elevator switch door detection method based on light stream Download PDFInfo
- Publication number
- CN106204659B CN106204659B CN201610590538.1A CN201610590538A CN106204659B CN 106204659 B CN106204659 B CN 106204659B CN 201610590538 A CN201610590538 A CN 201610590538A CN 106204659 B CN106204659 B CN 106204659B
- Authority
- CN
- China
- Prior art keywords
- door
- light stream
- mean value
- less
- information
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
-
- 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
Abstract
The elevator switch door detection method based on light stream that the present invention relates to a kind of.The present invention calculates Optic flow information in the area-of-interest of video, and extracts the Optic flow information mean value for moving downward and moving right respectively in area-of-interest, it is contemplated that the influence of illumination, while being also extracted valid pixel number;The Optic flow information and Pixel Information that extract are preserved, the Optic flow information of statistical number frame is also wanted simultaneously, in order to be avoided that the interference of illumination leads to the error in data of individual frames, by the analysis to light stream direction movable information, output respectively is the useful information for starting to open the door, open the door completely, start to close the door and close the door completely.The present invention has fully considered the convenience that scene is implemented, and is directly analyzed with the data of the installed camera acquisition of elevator;Without carrying out any change to scene, while the problem of blocked in view of personnel, only the upper part of elevator door is analyzed, therefore the elevator switch door information of robust can be obtained.
Description
Technical field
The invention belongs to intelligent monitoring technology fields, are related to a kind of elevator switch door detection method based on light stream.
Background technology
With the continuous social and economic development, great variety has occurred in people's lives environment, and more and more cells are write
The buildings such as word building are rised sheer from level ground, and in order to facilitate the discrepancy of people, the buildings such as many cells and office building are all configured with elevator, still
With the substantial increase of elevator usage amount, in recent years, the frequency that switch gate failure occurs for elevator is higher and higher.When elevator is opened
After shutdown failure, people is caused to be stranded elevator, it is most important that get in touch with the external world within the shortest time, seek to rescue.
And traditional elevator monitoring is a video recording and display function, when a failure occurs it, only grade trapped persons press Help Point,
Until monitored personnel have found just come to rescue;And when calling for help, may there are Help Point failure or monitoring personnel
Do not have situations such as on duty, can not being rescued at the first time for trapped person will be made, therefore when elevator switch door breaks down
When, so that monitoring personnel is quickly obtained information automatically, and become the emphasis studied at present to save trapped person.
Invention content
The elevator switch door detection method based on light stream that in view of the deficiencies of the prior art, the present invention provides a kind of.
The central scope of the present invention:Optic flow information is calculated in the area-of-interest of video, and is carried respectively in area-of-interest
The Optic flow information mean value for being orientated left movement and moving right, it is contemplated that the influence of illumination, while being also extracted valid pixel number;
The Optic flow information and Pixel Information that extract are preserved, while also wanting the Optic flow information of statistical number frame, in order to be avoided that
The interference of illumination leads to the error in data of individual frames, and by the analysis to light stream direction movable information, output is to start out respectively
Door opens the door, starts the useful information closed the door and closed the door completely completely.
The method of the present invention includes the following steps:
Step 1, to scene carry out divider calipers then, be arranged four effective regions, respectively left side side door, the right side door, in
Between paste each region in left and right of dangerous picture, the rule drawn in this way can be efficiently solved passed in and out personnel caused by the door
Block, and make algorithm generate result inaccuracy problem;The data volume of algorithm process is decreased simultaneously, it is effective to be promoted
The speed of service of algorithm.
Step 2, the movable information that respectively four effective coverages are carried out with horizontal direction count, i.e., are more than to moving displacement
0.1 pixel, respectively to move downward and move right pixel Optic flow information carry out it is cumulative and, while statistics move downward and
Move right pixel number it is cumulative and, then Optic flow information it is cumulative and divided by number of pixels it is cumulative and, obtain respectively to the left and
Move right the mean value of displacement.
Step 3, the displacement mean value to the left and to the right for four effective coverages being calculated acquisition are saved in mean value queue, queue
Length is 8.
Step 4, the mean value queue according to calculating, calculate separately the variance of the mean value queue.
Step 5, the mark value that four regions are obtained by the judgement of mean value and variance.
Each mark value is stored in respective mark queue by step 6, which is 16.
The number that step 7, the label that each mark queue is counted according to 16 systems are.
Step 8, basis are labeled as 1 number sequential value, judge the state status when front door.
Step 9, due to state change be start open the door->Open the door completely->Start shutdown->It closes the door and changes completely, so most
Afterwards according to status information, a state is determined;
After step 10, status information that must to the end and export.
Beneficial effects of the present invention:
1, with the method based on video monitoring, the switch door state of elevator is detected, the switch gate of elevator can be believed in real time
Breath feeds back to monitoring personnel.
2, in view of the influence of illumination, the light stream statistics that the inside adds valid pixel calculates mean value, and is based on mean value team
The variance of row calculates.
3, by the judgement to mean value and variance, can effectively overcome elevator nobody when turn off the light and when someone
Illuminance abrupt variation situation caused by turning on light;
4, changed by the continuous state of multiframe, keep switch gate state output more robust.
Description of the drawings
Fig. 1 is the method for the present invention flow chart.
Fig. 2 is the effective coverage in the present invention.
Specific implementation mode
The invention will be further described with attached drawing 1 with reference to embodiments:
Present invention is mainly applied to the detection of the switch gate of elevator, the switch door states for detecting elevator of energy robust, for event
When barrier occurs, it can quickly pinpoint the problems, improve efficiency, specific implementation step is as follows:
Step 1 carries out divider calipers then to the upper half department of scene, and four effective regions, respectively left side side door, the right side are arranged
Each region in left and right of avris door, the dangerous picture of intermediate patch, as shown in Figure 2.
Step 2 carries out peripheral maximal encasing box calculating to this four regions, obtains maximal encasing box coordinate, and extract most
Big bounding box image data, hereinafter referred to as image.
Step 3 carries out optical flow algorithm calculating to the image, obtains the Optic flow information of the image.
Step 4, the movable information that respectively four effective coverages are carried out with horizontal direction count, i.e., are more than to moving displacement
0.1 pixel, respectively to move downward and move right pixel Optic flow information carry out it is cumulative and, while statistics moves downward
With the number that moves right pixel it is cumulative and, then light stream it is cumulative and divided by number of pixels it is cumulative and, obtain respectively to the left and to
The mean value of right moving displacement.
Step 5, the displacement mean value to the left and to the right for four regions being calculated acquisition are saved in mean value queue, queue length
It is 8.
Step 6, the mean value queue according to calculating, calculate separately the mean variance of the mean value queue.
Step 7 assumes that the number of four regions from left to right is respectively:L1,L2,R1,R2.
If the light stream mean value that step 8, L2 are moved downward is more than 50 and is less than 250, while the light stream that R1 is moved right
Mean value is more than 50 and is less than 250, and mean variance is less than 2000;Start to open the door with regard to setting and be labeled as 1, is otherwise 0.
If the light stream mean value that L1 is moved downward is more than 50 and is less than 250, while the light stream mean value that R2 is moved right is big
In 50 and be less than 250, and mean variance be less than 2000;It just sets up standard-sized sheet door and is labeled as 1, be otherwise 0.
If the light stream mean value that L1 is moved right is more than 50 and is less than 250, while the light stream mean value that R2 is moved downward is big
In 50 and be less than 250, and mean variance be less than 2000;Start to close the door with regard to setting and be labeled as 1, is otherwise 0.
If the light stream mean value that L2 is moved right is more than 50 and is less than 250, while the light stream mean value that R1 is moved downward is big
In 50 and be less than 250, and mean variance be less than 2000;It just sets up fully closed door and is labeled as 1, be otherwise 0.
Each mark value is stored in respective mark queue by step 9, which is 16.
The number that step 10, the label that each mark queue is counted according to 16 systems are.
If step 11 starts the statistical value closed the door, it is label 1 continuous three frame occur all, and being judged as current state is
Start to close the door.
If the statistical value closed the door completely, it is label 1 continuous four frame occur all, behind four frame flags be 0, be judged as working as
Preceding state is to close the door completely.
If starting the statistical value to open the door, it is label 1 continuous three frame occur all, and it is to start out to be judged as current state
Door.
If the statistical value to open the door completely, it is label 1 continuous four frame occur all, behind four frame flags be 0, be judged as working as
Preceding state is to open the door completely.
Step 12, due to state change be start open the door->Open the door completely->Start shutdown->It closes the door and changes completely, so
Finally according to status information, a state is determined.
After step 13, status information that must to the end and export.
The present invention has fully considered the convenience that scene is implemented, directly use the data that the installed camera of elevator acquires into
Row analysis;Without carrying out any change to scene, while the problem of blocked in view of personnel, only the upper part of elevator door is carried out
Analysis, therefore the elevator switch door information of robust can be obtained.
The foregoing is only a preferred embodiment of the present invention, is not intended to limit the scope of the present invention, should band
Understand, the present invention is not limited to implementation as described herein, the purpose of these implementations description is to help this field
In technical staff put into practice the present invention.
Claims (5)
1. the elevator switch door detection method based on light stream, it is characterised in that this approach includes the following steps:
Step 1 carries out divider calipers then to the first half elevator switch door of scene, and four effective regions, respectively left side side are arranged
Each region in left and right of door, the right side door, the dangerous picture of intermediate patch, according to from left to right number is respectively:L1,L2,R1,
R2;
Step 2 carries out peripheral maximal encasing box calculating to this four regions, obtains maximal encasing box coordinate, and extract maximum packet
Enclose box image data;
Step 3 carries out optical flow algorithm calculating to the image, obtains the Optic flow information of the image;
Step 4, the movable information that respectively four effective coverages are carried out with horizontal direction count, and obtain to the left and move right displacement
Mean value;
Step 5, the displacement mean value to the left and to the right for four regions being calculated acquisition are saved in mean value queue;
Step 6, the mean value queue according to calculating, calculate separately the mean variance of the mean value queue;
If the light stream mean value that step 7, L2 are moved downward is more than 50 and is less than 250, while the light stream mean value that R1 is moved right
More than 50 and it is less than 250, and mean variance is less than 2000;Start to open the door with regard to setting and be labeled as 1, is otherwise 0;
If the light stream mean value that L1 is moved downward is more than 50 and is less than 250, while the light stream mean value that R2 is moved right is more than 50
And it is less than 250, and mean variance is less than 2000;It just sets up standard-sized sheet door and is labeled as 1, be otherwise 0;
If the light stream mean value that L1 is moved right is more than 50 and is less than 250, while the light stream mean value that R2 is moved downward is more than 50
And it is less than 250, and mean variance is less than 2000;Start to close the door with regard to setting and be labeled as 1, is otherwise 0;
If the light stream mean value that L2 is moved right is more than 50 and is less than 250, while the light stream mean value that R1 is moved downward is more than 50
And it is less than 250, and mean variance is less than 2000;It just sets up fully closed door and is labeled as 1, be otherwise 0;
Each mark value is stored in respective mark queue by step 9;
The number that step 10, the label that each mark queue is counted according to 16 systems are;
Step 11, basis are labeled as 1 number sequential value, judge the state status when front door;
Step 12, state change are to start enabling->Open the door completely->Start shutdown->It closes the door completely transformation, according to state status,
Determine a state;
Step 13, status information to the end and output.
2. the elevator switch door detection method according to claim 1 based on light stream, it is characterised in that:Step 4 is specifically:
It is more than the pixel of 0.1 pixel to moving displacement, tires out respectively to moving downward and moving right pixel Optic flow information
Adduction, at the same statistics move downward and move right pixel number it is cumulative and, then Optic flow information is cumulative and divided by pixel
Point number it is cumulative and, to obtain to the left and move right the mean value of displacement.
3. the elevator switch door detection method according to claim 1 based on light stream, it is characterised in that:The mean value team
Row length is 8.
4. the elevator switch door detection method according to claim 1 based on light stream, it is characterised in that:The label team
Row length is 16.
5. the elevator switch door detection method according to claim 1 based on light stream, it is characterised in that:Judge in step 11
When the state status at front door is specifically:
If starting the statistical value closed the door, it is label 1 continuous three frame occur all, and it is to start to close the door to be judged as current state;
If the statistical value closed the door completely, it is label 1 continuous four frame occur all, behind four frame flags be 0, be judged as current shape
State is to close the door completely;
If starting the statistical value to open the door, it is label 1 continuous three frame occur all, and it is to start to open the door to be judged as current state;
If the statistical value to open the door completely, it is label 1 continuous four frame occur all, behind four frame flags be 0, be judged as current shape
State is to open the door completely.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610590538.1A CN106204659B (en) | 2016-07-26 | 2016-07-26 | Elevator switch door detection method based on light stream |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610590538.1A CN106204659B (en) | 2016-07-26 | 2016-07-26 | Elevator switch door detection method based on light stream |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106204659A CN106204659A (en) | 2016-12-07 |
CN106204659B true CN106204659B (en) | 2018-11-02 |
Family
ID=57494981
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610590538.1A Active CN106204659B (en) | 2016-07-26 | 2016-07-26 | Elevator switch door detection method based on light stream |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106204659B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106909917A (en) * | 2017-03-21 | 2017-06-30 | 成都市极米科技有限公司 | A kind of elevator door state monitoring method and device |
CN107187980B (en) * | 2017-05-26 | 2019-03-15 | 杭州光珀智能科技有限公司 | The detection method and detection device of elevator door folding in a kind of lift appliance |
JP6673617B1 (en) * | 2019-03-20 | 2020-03-25 | 東芝エレベータ株式会社 | Image detection device |
CN113408352A (en) * | 2019-05-28 | 2021-09-17 | 浙江大华技术股份有限公司 | Pedestrian abnormal behavior detection method, image processing device and storage device |
CN112347862A (en) * | 2020-10-20 | 2021-02-09 | 浙江新再灵科技股份有限公司 | Elevator door fault real-time detection method based on machine vision |
CN112598746B (en) * | 2020-12-18 | 2022-03-11 | 北京声智科技有限公司 | Elevator door opening and closing detection method and device, readable medium and electronic equipment |
CN112723077A (en) * | 2021-01-12 | 2021-04-30 | 南通润雅机电科技有限公司 | Elevator door opening and closing detection method based on optical flow |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006276969A (en) * | 2005-03-28 | 2006-10-12 | Mitsubishi Electric Corp | Violence detecting device |
JP2008137755A (en) * | 2006-11-30 | 2008-06-19 | Hitachi Building Systems Co Ltd | Elevator operation control device |
CN102025976A (en) * | 2009-09-18 | 2011-04-20 | 东芝电梯株式会社 | Savage act detection device and savage act detection method |
CN102145839A (en) * | 2010-02-09 | 2011-08-10 | 东芝电梯株式会社 | Elevator remaining object detection device |
CN104156932A (en) * | 2013-05-13 | 2014-11-19 | 哈尔滨点石仿真科技有限公司 | Moving object segmentation method based on optical flow field clustering |
-
2016
- 2016-07-26 CN CN201610590538.1A patent/CN106204659B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006276969A (en) * | 2005-03-28 | 2006-10-12 | Mitsubishi Electric Corp | Violence detecting device |
JP2008137755A (en) * | 2006-11-30 | 2008-06-19 | Hitachi Building Systems Co Ltd | Elevator operation control device |
CN102025976A (en) * | 2009-09-18 | 2011-04-20 | 东芝电梯株式会社 | Savage act detection device and savage act detection method |
CN102145839A (en) * | 2010-02-09 | 2011-08-10 | 东芝电梯株式会社 | Elevator remaining object detection device |
CN104156932A (en) * | 2013-05-13 | 2014-11-19 | 哈尔滨点石仿真科技有限公司 | Moving object segmentation method based on optical flow field clustering |
Also Published As
Publication number | Publication date |
---|---|
CN106204659A (en) | 2016-12-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106204659B (en) | Elevator switch door detection method based on light stream | |
CN101795395B (en) | System and method for monitoring crowd situation | |
CN104123544B (en) | Anomaly detection method and system based on video analysis | |
CN103729858B (en) | A kind of video monitoring system is left over the detection method of article | |
CN112347873B (en) | Ladder control method | |
CN103617699A (en) | Intelligent safety monitor system of electric power working site | |
CN103108159A (en) | Electric power intelligent video analyzing and monitoring system and method | |
CN104470058A (en) | Partition illumination energy-saving switch control system and control method thereof | |
CN105096406A (en) | Video analysis system used for architectural energy consumption equipment and intelligent building management system | |
CN104050480A (en) | Cigarette smoke detection method based on computer vision | |
CN103400111A (en) | Method for detecting fire accident on expressway or in tunnel based on video detection technology | |
CN103236249B (en) | The method and apparatus that display device regulates automatically | |
CN103150736A (en) | Camera motion detecting method based on video monitoring | |
CN106034237A (en) | Mixed coding method and system based on coding switching | |
CN109583339A (en) | A kind of ATM video brainpower watch and control method based on image procossing | |
CN109858338A (en) | A kind of identification and crowd behaviour parser of crowd density estimation | |
CN114917519B (en) | Building intelligent fire control system | |
CN113111771A (en) | Method for identifying unsafe behaviors of power plant workers | |
CN109034038A (en) | A kind of fire identification device based on multi-feature fusion | |
CN206209759U (en) | A kind of online fugitive personnel's automatic identification synthesis police service booth | |
CN102609685B (en) | Shadowing judging method of image type fire detector | |
CN110809137A (en) | Campus safety trampling prevention monitoring system and method | |
CN110348327A (en) | Realize the method and device that Articles detecting is left in monitoring scene | |
CN114283544A (en) | Railway platform intrusion monitoring system and method based on artificial intelligence | |
CN105741503A (en) | Parking lot real time early warning method under present monitoring device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Elevator door opening and closing detection method based on optical flow Effective date of registration: 20190821 Granted publication date: 20181102 Pledgee: Hangzhou Yuhang Small and Medium-sized Enterprise Transfer Service Co., Ltd. Pledgor: ZHEJIANG ICARE VISION TECHNOLOGY CO., LTD. Registration number: Y2019330000020 |