A Driver Safety System Using Raspberry Pi and Computer Vision

<div class="section abstract"><div class="htmlview paragraph">Road safety remains a critical concern globally, with millions of lives lost annually due to road accidents. In India alone, the year 2021 witnessed over 4,12,432 road accidents resulting in 1,53,972 fatalities and 3,84,448 injuries. The age group most affected by these accidents is 18-45 years, constituting approximately 67% of total deaths. Factors such as speeding, distracted driving, and neglect to use safety gear increases the severity of these incidents. This paper presents a novel approach to address these challenges by introducing a driver safety system aimed at promoting good driving etiquette and mitigating distractions and fatigue. Leveraging Raspberry Pi and computer vision techniques, the system monitors driver behavior in real-time, including head position, eye blinks, mouth opening and closing, hand position, and internal audio levels to detect signs of distraction and drowsiness. The system operates in both passive and active modes, providing alerts and alarms to the driver while also implementing a negative reinforcement mechanism. Through a negative reinforcement system which consists of not starting the car if the driver is distracted and sleepy in the previous trip, hence discourages distracted or drowsy driving behavior. Various methods for detecting the driver drowsiness have been experimented with. The one with the highest accuracy were used in the system. By providing real-time feedback and implementing a proactive deterrent, the system aims to promote safer driving practices and contribute to enhancing road safety. Experimental results demonstrate the effectiveness of the proposed system in identifying and reducing driver distractions and fatigue, thus contributing to a reduction in the number of accidents.</div></div>