Efficient Train Driver Drowsiness Detection on Machine Vision Algorithms

In order to give a warning to the drowsy driver, this paper proposes a new Fatigue Driving Detection Algorithm. AdaBoost algorithm is applied to fast detect and track human faces, and the algorithm is implemented in FPGA; differential template-based multi-algorithms are used to localize human eyes and recognize eye states; PERCLOS algorithm is adopted to analyze and determine whether a person is fatigue. Test is implemented based on DM368 add FPGA platform ,and confirms that the algorithm used in the paper can quickly and accurately locate the face and the human eye, and determine the status of the driver's fatigue. Experimental results show that the algorithm has high recognition accuracy and robust performance under real train driving environment, even in head rotation, eye movements and wearing glasses was the case.