Detection Of Driver’s Fatigue Driving Based On Adaboost Algorithm

Since Ford invented the car, the car ownership is growing.the road has becomeincreasingly crowded, which produce more and more traffic accidents, most of whichare caused due to driver fatigue.the traffic accidents to families and society havebrought great harm and impact.developing a timely driver fatigue detection system withhigh accuracy is very necessary, but also will bring very significant social and economicbenefits.Based on this, from the face location, face tracking, eye contour detection andfatigue decision rule to start. Fatigue driving detection method based on Adaboostalgorithm is developed, the main work is as follows:1．Describe the basic face location method, Have a research on Adaboostclassification based on Haar features, and three typical adaboost algorithm error ratecomparison due to the increase of cascade number.2．Have a research on the compression sensing algorithm, based on the human facelocation,Use sparse matrix to extract facial feature information in the known position ofthe face, so as to create a face tracking model.3．Describe the human eye contour detection and template matching method, usean improved AAM algorithm to track the contour of the human eye, and explainscientific PERCLOS rule in current fatigue detection field,in order to achieve driverfatigue determination.4．The software platform development environment here is Microsoft VisualStudio2010. Combined with the open source OpenCV2.4.3algorithm library of theIntel Corporation.In order to make a fatigue detection interface, the entire systemachieve the detection of driver fatigue, and has a high accuracy rate, number of framesper second is about6, It meet the real-time requirement.