Fatigue State Detection for Tired Person in Presence of Driving Periods

Due to the increasing of traffic accidents, there is an urgent need control and reduce driving mistakes. Driver fatigue or drowsiness one these major Algorithms have been developed address this issue by detecting alerting driver dangerous condition. The problem algorithms their accuracy, well as time required detect status alert driver. accuracy represent a critical condition that affects reduction accidents. Several datasets used in development drowsy detection techniques. These data gathered from deriver's brain Electroencephalogram (EEG) signals video streaming recordings behavior. This paper develops two distinct approaches, first based on use machine learning classifiers second depends deep models produce high-performance system. approach process EEG signals, whereas streams. In classifiers, Support Vector Machine (SVM) provides up 98% which highest among other five deployed classifiers. models, Convolution Neural Network (CNN) 99% models. experimental results demonstrate proposed provide with shortest Testing Time (TT) when compared all recent efficient algorithms.