Open AccessOptimization of Fatigue Detection Method under Altitude Changes in Plateau Region Based on MTCNN
Author(s) -
Sipeng Han,
Jingyang Tan,
Qianzhi Jiao,
Bo Tang,
Yanqi Luo,
Xuguang Yang
Publication year - 2021
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/692/4/042016
Subject(s) - computer science , artificial intelligence , plateau (mathematics) , feature (linguistics) , filter (signal processing) , noise reduction , computer vision , pattern recognition (psychology) , mathematics , mathematical analysis , philosophy , linguistics
Fatigue driving is the main cause of traffic accidents, and research on fatigue driving detection algorithms is of great significance to improve road safety. This paper proposes an image processing method based on MTCNN model detection optimization, Perform median filter denoising before P-Net training to improve the detection rate of night faces, then, the ASM algorithm is used to detect the facial feature points, and finally the PERCLOS principle is used to analyze the driving fatigue state. The experimental results show that the method has a high detection rate, can be applied to fatigue detection at different altitudes, and has strong practicability.