CN106022235A - Missing child detection method based on human body detection - Google Patents
Missing child detection method based on human body detection Download PDFInfo
- Publication number
- CN106022235A CN106022235A CN201610314230.4A CN201610314230A CN106022235A CN 106022235 A CN106022235 A CN 106022235A CN 201610314230 A CN201610314230 A CN 201610314230A CN 106022235 A CN106022235 A CN 106022235A
- Authority
- CN
- China
- Prior art keywords
- rectangle frame
- infrared sensor
- child
- target
- search window
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 36
- 238000000034 method Methods 0.000 claims abstract description 23
- 238000004891 communication Methods 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 2
- 230000014759 maintenance of location Effects 0.000 claims description 2
- 238000003860 storage Methods 0.000 claims description 2
- 238000012360 testing method Methods 0.000 claims description 2
- 230000005611 electricity Effects 0.000 description 2
- 101000633607 Bos taurus Thrombospondin-2 Proteins 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005206 flow analysis Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
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
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Burglar Alarm Systems (AREA)
Abstract
The invention relates to a missing child detection method based on human body detection. In the case of very low power consumption, human body targets are initially detected using an infrared sensor. After a suspected target is found through initial detection, a camera is started, and whether there is a missing child is confirmed using a video target re-checking method. Through a double-detection mode, the power consumption of the system is reduced, the false alarm rate and missing alarm rate of missing child detection are reduced, and the reliability of missing child detection is improved.
Description
Technical field
The present invention relates to a kind of omission child's detection method, belong to technical field of video monitoring.
Background technology
In recent years, repeatedly occurred to omit child on school bus and child's death by accident of causing, caused the strong pass of society
Note.In order to take precautions against this type of accident, except to responsible person concerned and by bus child carry out regular safety education and supervision and check
Outside, in addition it is also necessary to improve discovery and the pre-alerting ability omitting child technically.Along with the fast development of computer vision technique,
Human Detection is more and more ripe, patent " in car, delay occupant state identifies and danger state control system " (numbering:
CN201510779628.0,2015) leave over by whether there are passenger on pressure transducer detection seat, this detection method
Cannot be distinguished by people and thing, also cannot detect corridor or other sensor region is not installed leave over child, accuracy of detection is also simultaneously
The highest.Document " Deeply Learned Attributes for Crowded Scene Understanding " (CVPR,
2015) use and whether degree of depth learning method detection image exists human body target, but the method blocks feelings for existing on vehicle
Child's Detection results of condition is bad.Document " (Automatic Head Detection for Passenger Flow
Analysis in Bus Surveillance Videos " (CISP, 2012) by car door set up video camera statistics come in and go out
Number, counting precision is the highest, it is impossible to accurately judge whether omit child on car.
Summary of the invention
Spy of the present invention proposes the reliable detection method of omission child of a kind of low-power consumption.
The technical scheme is that in the case of extremely low power dissipation, at the beginning of using infrared sensor that human body target is carried out
Inspection；After Preliminary detection to suspicious object, restart video camera, use video object to check method and be confirmed whether that existence is left over
Child, detects pattern by bilayer, both can reduce system power dissipation, can reduce again false alarm rate and the false dismissal omitting child's detection
Rate, thus improve the reliability leaving over child's detection.
For achieving the above object, the present invention uses following technical scheme, flow chart as shown in Figure 1:
1, infrared human body detection initial survey
After school bus stops, owing to electromotor does not works, the accumulator of school bus is used to detect when equipment is powered special to leaving over child
Do not need the power problems of attention location system.If the power consumption of system is too high, easily expends the too much electricity of accumulator, thus cause school bus
Cannot light a fire.
The child of leaving over of the present invention detects equipment and mainly includes that three module: ARM process plate, infrared sensor, take the photograph
Camera (band infrared LED).Wherein, ARM processes plate and infrared sensor power consumption is the lowest, and video camera power consumption is higher, if at light
Infrared LED is opened, then power consumption is the highest in the case of line is dark.In order to reduce system power dissipation, the present invention uses at ARM
Reason plate controls the power supply of each module, only processes plate to ARM and infrared sensor is powered, the merit of the two module under silent status
Consume the lowest.Only when infrared sensor detects suspicious object, just notice arm processor is on the big video camera of power consumption
Electricity.Concretely comprise the following steps:
Step 1:ARM processes plate inquiry and connects the GPIO level of infrared sensor；
Step 2: if the GPIO level connecting infrared sensor is high level, represents that infrared sensor detects target, then
Power on to video camera, close infrared sensor power supply；Otherwise, Step 1 is proceeded to.
Step 3:ARM processes plate startup video object and checks thread.
Step 4: inquire about video object in 3 minutes sections and check the testing result of thread, if video object is checked
Thread detects target, then start the audio alert of vehicle, and start image and information of vehicles in wireless communication module transmission car
To school bus director and supervisory department, after being sent, close wireless communication module, voice module.
Step5: close video camera, power on to infrared sensor simultaneously, proceed to Step 1.
2, video object is checked
For reducing the false-alarm that infrared sensor causes, the present invention, after infrared sensor triggers and reports to the police, uses video analysis side
Suspicious object is checked by method.Concretely comprise the following steps:
Step1: edge calculations
In view of certainly existing gray difference between target and background, therefore the present invention first uses gradient operator on gray level image
Ask for image border.PixelGradient can be expressed as
Wherein,f (x ,y) represent pixelThe gradation of image at place,WithRepresent its edge respectivelyWithDirection
Gradient.Concrete Grad can use image to represent with the convolution of gradient operator template, and the present invention uses Sobel operator,
As shown in Figure 2.
PixelGradient modulus value can be expressed as
For each pixel, if its gradient modulus value m is more than threshold value T1, then it is assumed that this pixel is edge pixel point.This
In invention, threshold value T1 takes empirical value 10.
Step2: target coarse positioning
Image is carried out multiple dimensioned search, specifically, if minimum search window size is(namely suspicious object
Minimum dimension, in the present invention, picture size is 640 × 480,It is taken as 10,It is taken as 20), maximum search window size is(namely the full-size of suspicious object, in the present inventionIt is taken as 60,It is taken as 120).Advanced every trade is searched
Rope, now the height of search window is constant, each take turns search complete (i.e. till the image upper left corner starts to arrive the lower right corner) it
Rear hatch width increases by 1, until it reaches maximized window width；Then search window height increases by 1, continues to repeat row and searches
Rope, until search window width and highly respectively reachingWith.In multiple dimensioned search procedure, it is judged that each is searched
Whether rope window meet following two condition:
(1) edge pixel point is comprised in search window；
(2) there is not edge pixel point on the four edges of search window, namely all edge pixels point is all inside search window.
If certain search window meets above-mentioned two condition, then it is assumed that this search window there may be target, store this square
Shape frame.
Then, the rectangle frame of all storages is merged operation, specifically, the rectangle frame that position exists overlap is closed
And, after merging, starting point coordinate and the size of rectangle frame take merging the first two rectangle frame coordinate and the meansigma methods of size, simultaneously for
Rectangle frame after merging, records its rectangle frame merged sum, as the score of this rectangle frame.
Finally, for the rectangle frame after merging, if its score is more than threshold value T2, then retain this rectangle frame, it is believed that this square
Suspicious object is comprised in shape frame；Otherwise, this rectangle frame is deleted.In the present invention, threshold value T2 takes empirical value 5.
Step3: Hough loop truss
Each rectangle frame that traversal Step2 detects, to each rectangle frame, scans edge pixel therein point, uses Hough
Circle detection method judges whether to exist in this rectangle frame approximate circle, and records round radius R.If T3 < R < T4, then it is assumed that exist
Leaving over child, video object checks result for there is child's target.If all rectangle frames the most do not exist leaves over child, then it is assumed that
Image does not has child's target.Wherein, T3 and T4 is empirical value, in the present invention T3=5, T4=30.
It is an advantage of the current invention that: use infrared human body detection initial survey and video object to check double-deck detection pattern, both may be used
To reduce system power dissipation, false alarm rate and the false dismissed rate omitting child's detection can be reduced again.
Accompanying drawing explanation
Fig. 1 leaves over child's overhaul flow chart；
Fig. 2 Sobel operator template.
Detailed description of the invention
The reliable detection method of omission child of a kind of low-power consumption, uses infrared sensor that human body target is carried out initial survey；?
Preliminary detection, to after suspicious object, restarts video camera, uses video object to check method and is confirmed whether that child is left in existence,
Detect pattern by bilayer, both can reduce system power dissipation, false alarm rate and the false dismissed rate omitting child's detection can be reduced again, from
And improve the reliability leaving over child's detection.
Claims (1)
1. omission child's detection method based on human body detection, it is characterised in that use infrared sensor that human body target is carried out
Initial survey；After Preliminary detection to suspicious object, restart video camera, use video object to check method and be confirmed whether to there is something lost
Stay child,
Flow process is as follows:
(1), infrared human body detection initial survey
The child that leaves over of the present invention detects equipment and includes that three module: ARM process plate, infrared sensor, video cameras, uses ARM
Process plate and control the power supply of each module, under silent status, only process plate to ARM and infrared sensor is powered, the two module
Power consumption is the lowest, and only when infrared sensor detects suspicious object, just notice arm processor powers on to video camera, concrete steps
For:
Step 1:ARM processes plate inquiry and connects the GPIO level of infrared sensor；
Step 2: if the GPIO level connecting infrared sensor is high level, represents that infrared sensor detects target, then
Power on to video camera, close infrared sensor power supply；Otherwise, Step 1 is proceeded to；
Step 3:ARM processes plate startup video object and checks thread；
Step 4: inquire about video object in 3 minutes sections and check the testing result of thread, if video object checks thread
Target detected, then start the audio alert of vehicle, and start wireless communication module and send in car image and information of vehicles to school
Car director and supervisory department, close wireless communication module, voice module after being sent；
Step5: close video camera, power on to infrared sensor simultaneously, proceed to Step 1；
(2), video object is checked
For reducing the false-alarm that causes of infrared sensor, after infrared sensor triggers and reports to the police, use video analysis method to can
Doubtful target is checked, and concretely comprises the following steps:
Step2.1: edge calculations
In view of certainly existing gray difference between target and background, the present invention first uses gradient operator to ask on gray level image
Image border, pixelGradient table be shown as
Wherein,f (x ,y) represent pixelThe gradation of image at place,WithRepresent its edge respectivelyWithDirection
Gradient, concrete Grad uses image to represent with the convolution of gradient operator template,
PixelGradient modulus value be expressed as
For each pixel, if its gradient modulus value m is more than threshold value T1, then it is assumed that this pixel is edge pixel point, this
In invention, threshold value T1 takes empirical value 10；
Step2.2: target coarse positioning
Image is carried out multiple dimensioned search, specifically, if minimum search window size is, namely suspicious object
Minimum dimension, in the present invention, picture size is 640 × 480,It is taken as 10,Being taken as 20, maximum search window size is, namely the full-size of suspicious object, in the present inventionIt is taken as 60,Being taken as 120, advanced every trade is searched
Rope, now the height of search window is constant, each take turns search complete, i.e. till the image upper left corner starts to arrive the lower right corner, it
Rear hatch width increases by 1, until it reaches maximized window width；Then search window height increases by 1, continues to repeat row and searches
Rope, until search window width and highly respectively reachingWith, in multiple dimensioned search procedure, it is judged that each is searched
Whether rope window meet following two condition:
Edge pixel point is comprised in (a) search window；
There is not edge pixel point on the four edges of (b) search window, namely all edge pixels point be all inside search window,
If certain search window meets above-mentioned two condition, then it is assumed that this search window there may be target, store this rectangle frame,
Then, the rectangle frame of all storages is merged operation, specifically, the rectangle frame that position exists overlap is merged,
After merging, starting point coordinate and the size of rectangle frame take merging the first two rectangle frame coordinate and the meansigma methods of size, simultaneously for conjunction
Rectangle frame after and, records its rectangle frame merged sum, as the score of this rectangle frame,
Finally, for the rectangle frame after merging, if its score is more than threshold value T2, then retain this rectangle frame, it is believed that this rectangle frame
Inside comprise suspicious object；Otherwise, deleting this rectangle frame, in the present invention, threshold value T2 takes empirical value 5；
Step2.3: Hough loop truss
Each rectangle frame that traversal Step2.2 detects, to each rectangle frame, scans edge pixel therein point, uses suddenly
Husband's circle detection method judges whether to exist in this rectangle frame approximate circle, and records round radius R, if T3 < R < T4, then it is assumed that would deposit
Leaving over child, video object checks result for there is child's target, if all rectangle frames the most do not exist leaves over child, then recognizes
For there is no child's target in image, wherein, T3 and T4 is empirical value, in the present invention T3=5, T4=30.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610314230.4A CN106022235B (en) | 2016-05-13 | 2016-05-13 | Missing child detection method based on human body detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610314230.4A CN106022235B (en) | 2016-05-13 | 2016-05-13 | Missing child detection method based on human body detection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106022235A true CN106022235A (en) | 2016-10-12 |
| CN106022235B CN106022235B (en) | 2021-05-28 |
Family
ID=57100206
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610314230.4A Expired - Fee Related CN106022235B (en) | 2016-05-13 | 2016-05-13 | Missing child detection method based on human body detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106022235B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108111821A (en) * | 2018-01-10 | 2018-06-01 | 深圳羚羊极速科技有限公司 | A kind of equipment for being integrally interconnected net video access gateway and edge calculations |
| CN109484292A (en) * | 2018-10-15 | 2019-03-19 | 上海理工大学 | The detection system and detection method of automatic detection school bus stayer |
| CN109558848A (en) * | 2018-11-30 | 2019-04-02 | 湖南华诺星空电子技术有限公司 | A kind of unmanned plane life detection method based on Multi-source Information Fusion |
| WO2019095887A1 (en) * | 2017-11-14 | 2019-05-23 | 苏州数言信息技术有限公司 | Method and system for realizing universal anti-forgotten sensing device for in-vehicle passengers |
| CN109927644A (en) * | 2017-12-15 | 2019-06-25 | 郑州宇通客车股份有限公司 | A kind of vehicle and child-resistant legacy device |
| CN110827317A (en) * | 2019-11-04 | 2020-02-21 | 西安邮电大学 | FPGA-based four-eye moving target detection and identification device and method |
| CN112805761A (en) * | 2018-09-27 | 2021-05-14 | 三菱电机株式会社 | Legacy detection device and legacy detection method |
| CN113361410A (en) * | 2021-06-07 | 2021-09-07 | 上海数川数据科技有限公司 | Child detection box filtering algorithm based on grid clustering |
| CN113997898A (en) * | 2021-11-30 | 2022-02-01 | 浙江极氪智能科技有限公司 | Living body detection method, apparatus, device and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1698381A (en) * | 2003-05-08 | 2005-11-16 | 西门子公司 | Method and device for detecting an object or a person |
| CN101520850A (en) * | 2009-04-17 | 2009-09-02 | 中国科学院计算技术研究所 | Construction method of object detection classifier, object detection method and corresponding system |
| US8886206B2 (en) * | 2009-05-01 | 2014-11-11 | Digimarc Corporation | Methods and systems for content processing |
-
2016
- 2016-05-13 CN CN201610314230.4A patent/CN106022235B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1698381A (en) * | 2003-05-08 | 2005-11-16 | 西门子公司 | Method and device for detecting an object or a person |
| CN101520850A (en) * | 2009-04-17 | 2009-09-02 | 中国科学院计算技术研究所 | Construction method of object detection classifier, object detection method and corresponding system |
| US8886206B2 (en) * | 2009-05-01 | 2014-11-11 | Digimarc Corporation | Methods and systems for content processing |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019095887A1 (en) * | 2017-11-14 | 2019-05-23 | 苏州数言信息技术有限公司 | Method and system for realizing universal anti-forgotten sensing device for in-vehicle passengers |
| CN109927644A (en) * | 2017-12-15 | 2019-06-25 | 郑州宇通客车股份有限公司 | A kind of vehicle and child-resistant legacy device |
| CN108111821A (en) * | 2018-01-10 | 2018-06-01 | 深圳羚羊极速科技有限公司 | A kind of equipment for being integrally interconnected net video access gateway and edge calculations |
| CN112805761A (en) * | 2018-09-27 | 2021-05-14 | 三菱电机株式会社 | Legacy detection device and legacy detection method |
| US11521478B2 (en) | 2018-09-27 | 2022-12-06 | Mitsubishi Electric Corporation | Left-behind detection device and left-behind detection method |
| CN112805761B (en) * | 2018-09-27 | 2024-03-08 | 三菱电机株式会社 | Legacy detection device and legacy detection method |
| CN109484292A (en) * | 2018-10-15 | 2019-03-19 | 上海理工大学 | The detection system and detection method of automatic detection school bus stayer |
| CN109484292B (en) * | 2018-10-15 | 2021-11-30 | 上海理工大学 | Detection system and detection method for automatically detecting passengers detained in school bus |
| CN109558848A (en) * | 2018-11-30 | 2019-04-02 | 湖南华诺星空电子技术有限公司 | A kind of unmanned plane life detection method based on Multi-source Information Fusion |
| CN110827317A (en) * | 2019-11-04 | 2020-02-21 | 西安邮电大学 | FPGA-based four-eye moving target detection and identification device and method |
| CN113361410A (en) * | 2021-06-07 | 2021-09-07 | 上海数川数据科技有限公司 | Child detection box filtering algorithm based on grid clustering |
| CN113361410B (en) * | 2021-06-07 | 2022-08-26 | 上海数川数据科技有限公司 | Child detection box filtering algorithm based on grid clustering |
| CN113997898A (en) * | 2021-11-30 | 2022-02-01 | 浙江极氪智能科技有限公司 | Living body detection method, apparatus, device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106022235B (en) | 2021-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106022235A (en) | Missing child detection method based on human body detection | |
| US11521497B2 (en) | Method and system for recognition of objects near ship by using deep neural network | |
| US10115029B1 (en) | Automobile video camera for the detection of children, people or pets left in a vehicle | |
| CN103839346B (en) | A kind of intelligent door and window anti-intrusion device and system, intelligent access control system | |
| JP2022118730A (en) | Vehicle door lock release method and device, system, vehicle, electronic apparatus and storage medium | |
| WO2019178851A1 (en) | Deep learning-based manhole cover loss detection system and method | |
| CN103366506A (en) | Device and method for automatically monitoring telephone call behavior of driver when driving | |
| US11265508B2 (en) | Recording control device, recording control system, recording control method, and recording control program | |
| CN105447459A (en) | Unmanned plane automation detection target and tracking method | |
| CN103839373A (en) | Sudden abnormal event intelligent identification alarm device and system | |
| CN109532665B (en) | Anti-suffocation system for vehicle and method thereof | |
| CN103400430A (en) | Metro door safety monitoring system based on video light curtain | |
| CN105809890A (en) | School-bus-safety-oriented missed-child detecting method | |
| CN106926794A (en) | Vehicle monitoring system and method thereof | |
| JP6140436B2 (en) | Shooting system | |
| CN108960131A (en) | The anti-people of mechanical garage is strayed into detection method | |
| KR101415848B1 (en) | Monitering apparatus of school-zone using detection of human body and vehicle | |
| KR101552564B1 (en) | Fusion security system based on gas sensor and IP network camera | |
| CN112784759A (en) | Elevator human detection identification method based on artificial intelligence similarity comparison | |
| TWI476735B (en) | Abnormal classification detection method for a video camera and a monitering host with video image abnormal detection | |
| Al Jarouf et al. | A hybrid method to detect and verify vehicle crash with haar-like features and svm over the web | |
| CN114373303A (en) | Mobile modularized public security intelligent inspection station system | |
| KR102107298B1 (en) | Image analysis apparatus and method | |
| TW201447825A (en) | Surveillance system for image recognition | |
| CN111564014A (en) | Shipborne detection terminal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210528 |