Driver Drowsiness and Distraction Detection: An Image Processing-Based Comparative Analysis for Improved Accuracy

Abstract

This research presents a comprehensive examination and implementation of driver drowsiness, distraction, and detection systems utilizing advanced image processing techniques. The literature review encompasses an in-depth analysis of drowsiness, distraction, and detection parameters, presented in tabulated form. The proposed architecture is detailed through flow charts outlining both software and hardware components. A comparative analysis of key parameters, along with their corresponding accuracy percentages, is provided in a structured table. The findings demonstrate that the proposed system exhibits superior accuracy compared to existing results. Through practical implementation, the system proves effective in accurately detecting driver sleepiness, classifying states as Sleepy, Drowsy, or Active. Notably, the proposed work achieves high accuracy, with eye detection accuracy at 98% and drowsiness accuracy at 96%, showcasing an improvement of approximately 10% when compared to existing solutions.