A Portable Fuzzy Driver Drowsiness Estimation System.

摘要: The adequate automatic detection of driver fatigue is a very valuable approach for the prevention traffic accidents. Devices that can determine drowsiness conditions accurately must inherently be portable, adaptable to different vehicles and drivers, robust such as illumination changes or visual occlusion. With advent new generation computationally powerful embedded systems Raspberry Pi, category real-time low-cost portable could become standard tools. Usually, proposed solutions using this platform are limited definition thresholds some defined indicator application expensive classification models limits their use in real-time. In research, we propose development low-cost, accurate, recognition device. device combines complementary measures derived from temporal window eyes (PERCLOS, ECD) mouth (AOT) states through fuzzy inference system deployed Pi with capability response. provides three degrees (Low-Normal State, Medium-Drowsy High-Severe Drowsiness State), was assessed terms its computational performance efficiency, resulting significant accuracy 95.5% state demonstrates feasibility approach.