Research On Tractor Driver Fatigue Detection Method Based On Machine Vision

The rapid development of China's agricultural economy has put forward higher requirements for agricultural technical equipment,especially the power system of the whole equipment--tractor.In order to meet the development demand of agricultural production,agricultural tractors are developing in the direction of large-scale,intelligent and high-tech.Although the driving environment of tractors has been improved in recent years,compared with other motor vehicle drivers,tractor drivers are greatly affected by noise,field environment,uneven illumination,poor driving environment,concentrated operation time,long driving time and other problems,which are easy to lead to driving fatigue.In this paper,the effectiveness of each method and its characteristics was tested by combining with the experimental conditions in this paper.Some algorithms were improved based on the detection principle of PERCLOS,and then the fatigue degree of tractor drivers was evaluated by using the detection principle assisted by blinking frequency.In this paper,the tractor driver's face is firstly extracted and detected.Gabor filtering method is adopted to preprocess the uneven illumination images.YCbCr color space is used for skin color segmentation.Secondly,on the basis of face location,human eye location features are used to roughly locate the human eye region.The connected region method is used to calculate the size of each connected region in the face image and remove the noise of human eye image(such as nose and hair,etc.).Finally,horizontal and vertical integral projections are used to locate the human eye region.The height to width ratio of human eyes is used to determine the closure of human eyes.Finally,the P80 standard of PERCLOS principle is used to judge the fatigue state of human eyes.If F value is greater than 0.16,the tractor driver is in the state of fatigue driving.