The Research Of Driver ’s Fatigue Monitoring Based On Machine Vision

According to the statistical data of traffic accidents in recent years,the main cause of traffic accidents is drivers of fatigue,which occurred on the highway driving is most serious,so the research of driver fatigue monitoring based on machine vision has important significance.First of all,a new method of rough positioning of driver’s eye is introduced.The infrared light source lighting,the image can avoid the interference of ambient light;then according to the shape feature of the human eye,a special vertical gradient operator is defined,which is convolution with the image pixel gray value,to eliminate interference;establish a new eye positioning template,used to calculate the sense of gradient evaluation within a region of interest value.When the gradient evaluation value reaches the maximum,the human eye can be located.In order to accelerate the speed of sliding matching,a method of segment matching is adopted.In this paper,the performance of the evaluation function is analyzed,and it shows that the evaluation function is adaptable in a certain range.At the same time,the influence of face rotation and face size on the value of evaluation function is analyzed.Then,the human eye is precisely located.Using the image preprocessing of the closed operation,combined with some eye area restrictions,a region labeling algorithm is used to mark the upper part of the template,the two areas closest to the eyebrows are the area to be sought.Finally,fatigue detection is carried out according to the shape characteristics of eyes.The main method is to measure the height of the upper and lower eyelids,and then select the P60 standard to calculate the eye PERCLOS values.The experimental results show that the method of this paper is faster and more accurate.There is no harsh requirement for the driver’s stance in the car,the positioning accuracy is relatively high;and the PERCLOS value is used to judge the fatigue state,which is accurate and reliable.