Heterogeneous Signal Fusion Method in Driving Fatigue Detection Signals

Driving fatigue is a physiological phenomenon that often occurs during driving. When the driver enters a fatigue state, they will become distracted and unresponsive, which can easily lead to traffic accidents. The driving fatigue detection method based on a single information source has poor stability in a specific driving environment and has great limitations. This work helps with being able to judge the fatigue state of the driver more comprehensively and achieving a higher accuracy rate of driving fatigue detection. This work mainly introduces research into different signal fusion methods to detect fatigue drive. This work will take the normal driver’s breathing signal, eye signals, and steering wheel signal as research objects and collect and isolate the characteristics of the fatigue detection signal. Research was then conducted on different signal fusion methods for the detected depth of breath. Change of steering angle, eyelid closure, and blinking marks and the fatigue driving experiment was designed to evaluate the results of different data fusion methods. Experimental results show that the detection accuracy of the heterogeneous signal fusion method in fatigue detection is as high as 80%.