AI Based Safety Alert for Car Door Opening Using Yolov5l Algorithm

Abstract

Vehicle accidents can occur due to various causes such as human error, mechanical failures, and adverse road conditions. These accidents can result in property damage, injuries, and even loss of life. Vehicle door collision detection and avoidance using ADAS (Advanced Driver Assistance System) is vital for autonomous driving. Accidents of this sort typically happen after the driver has parked the car and tries to open the door to get out, colliding with the vehicles coming up from behind, which may result in injuries to the commuters and the vehicle. To overcome this issue, the AI-SADOM (Artificial Intelligence – Safety Alert for Car Door Opening )  model uses YOLOv5l (You Only Look Once ) algorithm which gets the input of the safety distance of the vehicle approaching from the rear and then the YOLOv5l detects the presence of an object and calculates the distance. When a vehicle approaches from the rear, the SADOM system is designed to activate the dependable door control, and it will successfully apply vibration at the car's door that commuters may feel.