Detecting driver drowsiness using total pixel algorithm

Advancement in transportation technology certainly comes with numerous positive impacts. Nonetheless, some negative aspects including growing numbers of traffic accidents cannot be taken for granted. Factors that trigger traffic accidents range from human errors, vehicle mishaps, to the environment itself. Human error is somehow the factor that often causes traffic accidents. This research aims to propose a method of detecting drowsiness using the total pixel algorithm for drivers, with the help of video cameras connected to a computer. It was expected that it would help reduce the number of traffic accidents. The method employed in this research is detecting drivers' faces by segmenting RGB images into YCbCr color spectrum, determining the area of the eyes, and classifying eyes condition using total pixel algorithm. The system developed has been able to detect drowsiness in drivers without glasses with 90.5% to 92% accuracy. However, for the detection of objects with glasses ranging from 72.8% to 74.8% accuracy.