Automatic Braking System for Two Wheeler with Object Detection and Depth Perception

A tremendous growth in motorcycle traffic has occurred during the last decades in most developing countries and economically weak nations of the world. Road safety is a challenging aspect for two-wheeler users. In the face of this unfortunate aspect of motorcycle riding, it seems necessary to fully understand the dangers that users face, when traveling by two-wheeler and especially in an urban area. The goal of this paper is to implement an automated braking system. A camera is used to record the view in front of the user. Further, it is processed by using various machine learning algorithms to get the desired input for a brake control system. Algorithms used in this paper to process the images and videos are the YOLO object detection model and the mono depth model. They are used to detect the relative distance between the user and other vehicles and assist the user to apply the brake if there is a possibility of collision or accident.