CN107169441A - A kind of fast human-eye detection method - Google Patents
A kind of fast human-eye detection method Download PDFInfo
- Publication number
- CN107169441A CN107169441A CN201710328108.7A CN201710328108A CN107169441A CN 107169441 A CN107169441 A CN 107169441A CN 201710328108 A CN201710328108 A CN 201710328108A CN 107169441 A CN107169441 A CN 107169441A
- Authority
- CN
- China
- Prior art keywords
- frame
- human eye
- detected
- detection
- adaboost
- 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.)
- Withdrawn
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
- G06V40/18—Eye characteristics, e.g. of the iris
- G06V40/197—Matching; Classification
Abstract
The invention discloses a kind of fast human-eye detection method, comprise the following steps：S1：Human eye video image is gathered, present frame image to be detected is pre-processed；S2：If former frame ins detect human face region, it is tracked using difference method, is otherwise detected using AdaBoost, obtain FROI；S3：AdaBoost human eye detections are carried out to FROI, if detected successfully, human eye result are obtained, updated simultaneously；Otherwise human eye result is obtained using deformable matching detection, when AdaBoost methods it is continuous unsuccessfully more than 5 times when, abandon the detection to this frame, human eye detection failure, next frame is since S1；S4：Detected with multiple features grader；If by grader, S3 is directly carried out when lower frame is detected, otherwise, next frame is since S1, the detection failure of this frame.
Description
Technical field
Present invention relates particularly to a kind of fast human-eye detection method.
Background technology
Research shows both at home and abroad, and driver is in the state of fatigue, the perception to surrounding environment, dangerous judgement energy
Power and the manipulation ability of vehicle can all have different degrees of decline compared with normal condition, so as to cause traffic accident.Therefore, fatigue is driven
Sailing detection has particularly important meaning.
Current fatigue detecting system mainly uses the non-contact detection method based on image processing techniques, this fatigue driving
System is mainly detection and the state of analysis driver's eyes.When fatigue driving occurs, driver's frequency of wink can rise, and close
Increase between at the moment, the fatigue driving standard PERCLOS based on eye state is presently the most accurate while being also widely accepted
Standard.
The algorithm for being currently based on the fatigue driving detecting system of image processing techniques is mainly used based on Haar features
AdaBoost algorithms, this method determines human face region using cascade classifier first, then determines position of human eye with same method,
Judge whether fatigue driving behavior occurs by analyzing human eye state combination PERCLOS standards.And in practice environment,
Due to factors, AdaBoost algorithm tables such as the change of camera and face angulation, the rocking of driver head, vehicle shakes
Reveal poor robustness, refusing sincere height causes discrimination relatively low；Also there is the side using template matches during search human eye
Method, template matches calculation scale is larger, and when human eye is opened and closed, due to the change of target shape, using fixed single template
Target can be lost, amount of calculation is excessive if taking various template to match, it is impossible to reach requirement of real-time.
The content of the invention
The technical problem to be solved in the present invention is to provide a kind of fast human-eye detection method.
A kind of fast human-eye detection method, comprises the following steps：
S1：Human eye video image is gathered, if this frame image to be detected Img, last AdaBoost human eye detections successful result
AdaEyeImg, is pre-processed to present frame image to be detected Img；
S2：If former frame ins detect human face region, it is tracked, is otherwise carried out using AdaBoost using difference method
Detection, obtains FROI；
S3：AdaBoost human eye detections are carried out to FROI, if detected successfully, human eye result newEyeImg are obtained, updated simultaneously
AdaEyeImg；Otherwise human eye result newEyeImg is obtained using deformable matching detection, when AdaBoost methods continuously fail
During more than 5 times, the detection to this frame is abandoned, human eye detection fails, and next frame is since S1；
S4：By newEyeImg and AdaEyeImg, as input, detected with multiple features grader；If by grader,
NewEyeImg peripheral regions are directly then subjected to S3, otherwise, next frame is opened from S1 as next frame FROI when lower frame is detected
Begin, the detection failure of this frame.
The beneficial effects of the invention are as follows：
The present invention carries out human eye detection, and improve " people on the basis of AdaBoost algorithms using deformable matching method
Face-human-eye model ", selectively skips Face datection step, so as to be quickly and accurately positioned using the grader of multiple features
Human eye.
Embodiment
The present invention is further elaborated for specific examples below, but not as a limitation of the invention.
A kind of fast human-eye detection method, comprises the following steps：
S1：Human eye video image is gathered, if this frame image to be detected Img, last AdaBoost human eye detections successful result
AdaEyeImg, is pre-processed to present frame image to be detected Img；
S2：If former frame ins detect human face region, it is tracked, is otherwise carried out using AdaBoost using difference method
Detection, obtains FROI；
S3：AdaBoost human eye detections are carried out to FROI, if detected successfully, human eye result newEyeImg are obtained, updated simultaneously
AdaEyeImg；Otherwise human eye result newEyeImg is obtained using deformable matching detection, when AdaBoost methods continuously fail
During more than 5 times, the detection to this frame is abandoned, human eye detection fails, and next frame is since S1；
S4：By newEyeImg and AdaEyeImg, as input, detected with multiple features grader；If by grader,
NewEyeImg peripheral regions are directly then subjected to S3, otherwise, next frame is opened from S1 as next frame FROI when lower frame is detected
Begin, the detection failure of this frame.
Claims (1)
1. a kind of fast human-eye detection method, it is characterised in that comprise the following steps：
S1：Human eye video image is gathered, if this frame image to be detected Img, last AdaBoost human eye detections successful result
AdaEyeImg, is pre-processed to present frame image to be detected Img；
S2：If former frame ins detect human face region, it is tracked, is otherwise carried out using AdaBoost using difference method
Detection, obtains FROI；
S3：AdaBoost human eye detections are carried out to FROI, if detected successfully, human eye result newEyeImg are obtained, updated simultaneously
AdaEyeImg；Otherwise human eye result newEyeImg is obtained using deformable matching detection, when AdaBoost methods continuously fail
During more than 5 times, the detection to this frame is abandoned, human eye detection fails, and next frame is since S1；
S4：By newEyeImg and AdaEyeImg, as input, detected with multiple features grader；If by grader,
NewEyeImg peripheral regions are directly then subjected to S3, otherwise, next frame is opened from S1 as next frame FROI when lower frame is detected
Begin, the detection failure of this frame.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710328108.7A CN107169441A (en) | 2017-05-11 | 2017-05-11 | A kind of fast human-eye detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710328108.7A CN107169441A (en) | 2017-05-11 | 2017-05-11 | A kind of fast human-eye detection method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107169441A true CN107169441A (en) | 2017-09-15 |
Family
ID=59814872
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710328108.7A Withdrawn CN107169441A (en) | 2017-05-11 | 2017-05-11 | A kind of fast human-eye detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107169441A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1731418A (en) * | 2005-08-19 | 2006-02-08 | 清华大学 | Method of robust accurate eye positioning in complicated background image |
| CN101950355A (en) * | 2010-09-08 | 2011-01-19 | 中国人民解放军国防科学技术大学 | Method for detecting fatigue state of driver based on digital video |
-
2017
- 2017-05-11 CN CN201710328108.7A patent/CN107169441A/en not_active Withdrawn
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1731418A (en) * | 2005-08-19 | 2006-02-08 | 清华大学 | Method of robust accurate eye positioning in complicated background image |
| CN101950355A (en) * | 2010-09-08 | 2011-01-19 | 中国人民解放军国防科学技术大学 | Method for detecting fatigue state of driver based on digital video |
Non-Patent Citations (1)
| Title |
|---|
| 苏起扬 等: ""一种疲劳驾驶检测系统中快速人眼检测方法"", 《现代电子技术》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104732251B (en) | A kind of trainman's driving condition detection method based on video | |
| Yan et al. | Real-time driver drowsiness detection system based on PERCLOS and grayscale image processing | |
| CN104616438B (en) | A kind of motion detection method of yawning for fatigue driving detection | |
| CN102841354B (en) | Vision protection implementation method of electronic equipment with display screen | |
| CN103714660B (en) | System for achieving fatigue driving judgment on basis of image processing and fusion between heart rate characteristic and expression characteristic | |
| Nguyen et al. | Eye tracking system to detect driver drowsiness | |
| WO2016038784A1 (en) | Driver state determination apparatus | |
| US7620216B2 (en) | Method of tracking a human eye in a video image | |
| Xie et al. | Driver fatigue detection based on head gesture and PERCLOS | |
| Teyeb et al. | A drowsy driver detection system based on a new method of head posture estimation | |
| CN105117681A (en) | Multi-characteristic fatigue real-time detection method based on Android | |
| CN104361332A (en) | Human face eye region positioning method for fatigue driving detection | |
| CN103150870A (en) | Train motorman fatigue detecting method based on videos | |
| Punitha et al. | Driver fatigue monitoring system based on eye state analysis | |
| CN106203293A (en) | A kind of method and apparatus detecting fatigue driving | |
| Khan et al. | Real time eyes tracking and classification for driver fatigue detection | |
| Cai et al. | Gaze estimation driven solution for interacting children with ASD | |
| Long et al. | Near infrared face image quality assessment system of video sequences | |
| CN107169441A (en) | A kind of fast human-eye detection method | |
| TWI550440B (en) | Method and system for detecting person to use handheld apparatus | |
| Kumar | Morphology based facial feature extraction and facial expression recognition for driver vigilance | |
| KR20160115119A (en) | System and the method for recognizing drowsiness | |
| JPH0793700A (en) | Driver state monitoring device | |
| Kumagai et al. | Improving Accuracy of Traffic Sign Detection Using Learning Method | |
| Sooksatra et al. | A drowsiness detection method based on displacement and gradient vectors |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20170915 |