%0 Journal Article
%A HAN Tong-tong
%A SHANG Xue-da
%A WANG Xiao-yu
%T Driving Fatigue State Detecting Method Based on Densely Connected Convolutional Network
%D 2021
%R 10.13190/j.jbupt.2019-263
%J Journal of Beijing University of Posts and Telecommunications
%P 52-58
%V 44
%N 1
%X In order to solve the problem of traffic accidents caused by fatigue driving,a driving fatigue state detection method based on densely connected convolutional networks is proposed. Firstly,a camera is used to collect driving state video,the video frame image is obtained. The image processing technology is used to process images. Then the adaptive Boosting algorithm is used to detect the face,and then the gray integral projection and radial symmetry transformation algorithm are used to locate the driver's eye area. Moreover,the eye state is accurately distinguished through the densely connected network. Three densely connected blocks are set in the network to reduce feature parameters and speed up training. The sparse structure is used to reduce storage and enhance feature propagation. Finally,with the help of two fatigue parameters comprehensively judge the fatigue state of the driver,making the detection result more accurate. It is proved by a large number of qualitative and quantitative experiments that this method is better than the existing technology in terms of accuracy.
%U https://journal.bupt.edu.cn/EN/10.13190/j.jbupt.2019-263