Based On The Face Characteristics Of The Driver Fatigue Arithmetic Application Research

Due to the development of booming economy in recent years, automobile has been an essential part of our daily life. However, with the car ownership in China rising rapidly, the probability of traffic accidents also grows synchronously. In fact a survey suggest that fatigue driving is a new factor causing automobile accidents that can’t be ignored. So it’s significant for driver’s life safety to avoid the incidence of traffic accidents. So far the issue about how to prevent drivers from fatigue hasn’t been solved completely yet actually it is still far from implementation in real life.The most obvious characteristic of fatigue driving is the abnormal clothing of eyes and mouth. Moreover the face will also have a change at the same time. Thus when given a fatigue driving test for drivers, it is essential to test their main facial features, especially the features of eyes and mouth. And according to the tests, it should be effective to judge the state of fatigue.This paper is a study and solutions about how to test fatigue driving. Owing to the study, we can draw a conclusion that there are three main aspects of fatigue driving detection face detection, eyes detection as well as mouth detection. This paper aimed at color image take YCrCb color space method of skin color segmentation. Facial location based on skin color model for image preprocesing. Then reuse binarization and Otsu image finish segmentation facial region set up. The purpose is to get the face positioning.In the feature extraction of drivers’ eyes again, initial position location to the eyes. Finally determine the position of the eyes.This paper also studied the mouth. Driver’s mouth may yawn and speaking are influential to check. We based on k-means clustering and fisher linear classifier fusion algorithm of two kinds of methods to get the final localization in the mouth.The focus is on the driver fatigue detection, this article uses the detection method that is PERCLOS algorithm. The main characteristic of driver fatigue is dozing, so we use facial features detection algorithm for testing fatigue to finishing.Owing to the study, we can draw a conclusion that there are three main aspects of fatigue driving detection, face detection, eyes detection as well as mouth detection. For the driver’s fatigue situation rapidly and exactly can get the result of the driver fatigue.