Driver Distraction, alcohol and Obstacle Detection through Machine Learning: A Review

Driving is an inherently dangerous proposition as the vehicles are travelling at great speed which could lead to any minor inconsistencies or errors by the driver can lead to catastrophic results. To reduce such occurrences and provide a safe transport and travel for the users, there are carefully crafted rules and regulations that need to be abided. These rules are enforced by the traffic police and other regulatory authorities. But most of the time, the sheer number of vehicles on the road can overwhelm the authorities in their quest for compliance of the regulations. This leads to drunken driving and lethargic driver that is attempting to drive in that inebriated condition. This leads to unsafe conditions on the road that can lead to a mishap. There have been a multitude of approaches that are utilized for enabling the detection of drunkenness and distraction, but most of the approaches are either inaccurate or are highly intrusive. Therefore, this research proposes an effective technique for driver distraction along with alcohol and obstacle detection. The methodology employs the use of Region of Interest (ROI) in conjunction with Convolutional Neural Networks and Decision tree to provide highly accurate detection. This approach will be effectively outlined in the upcoming researches. Citation: Rupali Parte, Rohit Sangamnerkar, Kunal Patil, Ajay Kulkarni, Shubham Wadekar. "Driver Distraction, alcohol and Obstacle Detection through Machine Learning: A Review ." Global Research and Development Journal For Engineering 6.5 (2021): 48 - 52.