Driver Fatigue Recognition Based On AdaBoost Algorithm

The driver fatigue monitoring system is a technology which analyzes the face image by the computer,withdraws the effective characteristic information,and then distinguishes the specific condition of driver.Along with the rapid development of computer technology,the human face detection technology already applied to many domains.This paper mainly research the method that how to apply the human face detection technology in the human face sleepiness montior.According to the montior result,the system gives the state of fatigue early,warns promptly,and reduces the probability of the traffic accident caused by weariness.After research about the human face detection,camera controlling attention detection,this thesis bring forward a human eyes states detection algorithm based on AdaBoost algorithm, And implement a real-time attention detection system based on this algorithm.The main work done by this thesis includes three parts:1. Real-time human face detection system:This thesis discusses several typical human face detecting algorithms,and analyses and compares the adavntages and disadvantages between AdaBoost algorithm and skin color model algorithm.As a result,we choose AdaBoost algorithm to be the base algorithm for the human face detecting module of our drivers'attention detecting syetem.2. Based on the AdaBoost algorithm,this thesis analyses,compares and trains three kinds of different human eyes detectors,and as the result,we choose the best detector for fatigue detecting module of our drivers'attention detecting syestem.3. Based on what we have researched,we implement a real-time driver fatigue detecting syetem.The system contains two modules such as human face detecting module and driver attention detecting module.Based on AdaBoost algrithm, this thesis analyses,compares and trains the best blink eyes detector.Through this system,we can initially achieve the effective detection of driver attention and feed back to the server.