Research And Implementation Of Driver Fatigue Detection Algorithm Based On Facial Features

With the rapid development of economy,the number of car ownership has increased rapidly,which brings great convenience to people’s life,and the traffic safety problem is becoming more and more serious.Among the driver factors that cause traffic accidents,fatigue driving is one of the main reasons.Therefore,the study of the system for fatigue testing has important practical significance,which can effectively reduce traffic accidents caused by fatigue driving and improve traffic safety.Now,due to the relatively mature fatigue testing technology,the fatigue detection system performance improvement mainly depends on the detection algorithm improvement and innovation.In order to solve the problem of illumination variation and the limitation of single feature on the performance of fatigue detection system,based on the principle of infrared image and image processing technology,this paper focuses on the research of target feature extraction algorithm and multi-feature fusion algorithm,combining machine vision,machine learning,pattern recognition and other related knowledge.The driver fatigue detection algorithm based on multi-feature of human face is designed to extract and fuse multi-feature parameters such as eye closure degree,blink frequency and mouth-yawn in different lighting conditions.For the feature of eyes,a method was proposed.On the basis of the Adaboost algorithm,the eye region was enhanced by Retinex image,and the eyes closure and blink frequency were calculated by grid method.Then the eye features parameter was successfully achieved.For the feature of mouth,the mouth position was realized according to human eye coordinate.The feature parameters of mouth were extracted by connected domain method.In order to compensate for the limitation of single feature and further improve the reliability of the system,Bayesian conditional probability formula is used to fuse the eye closure degree,blink frequency and mouth yawn,and make the decision of fatigue state with PMERCLOS parameter.Finally,this paper focuses on the transplanting and training process of Adaboost algorithm.The driver fatigue detection algorithm based on multi-feature of human face is implemented and optimized on DM642 platform.The test results verify the adaptability,high real-time and accuracy of the algorithm.It is shown that the multi-feature fatigue detection system based on human face is of high scientific nature and validity,which is very important for driver fatigue detection.