A study on Lane Departure Warning System and Object Detection on Roads for Forward Collision Avoidance

Abstract: There are thousands of car accidents in the every year across the world. These accidents claim many lives and lots of properties. There are many different types of accidents, including rear end collisions, side swipes, head on collisions, collisions with stationary objects, accidents while changing lanes, and driving off the road. Seat belt usage, air bags, cruise control, rumble strips, and stricter vehicle safety requirements have all helped to reduce the number of accidents and fatalities. However, it is now possible to do more, by using intelligent driver assistant systems. In this paper, we propose a design lane departure warning system with object classification on road so that the driver can take smart and corrective measures to avoid the deviation from the lane as well as to avoid collision with the vehicle ahead of it. This safety system of Advanced Driver’s Assistance System (ADAS) can help in checking the road accidents. The system highly relies on the Open CV for lane line detection as Lane Departure Warning system (LDWS). The second part of the paper is object detection at the front of the vehicle for collision avoidance. It uses stereo camera as vision sensor. The real image obtained from vision system gives more accurate information to detect the target vehicle and gives satisfactory result. The output from the stereo vision sensor is fed to the controlling unit. Yolo (You Only Look Once) based real time object detection is used to determine the presence of the vehicles in the frame. Depth map for the vehicles located in the vision frame is obtained with help of stereo camera System Development Kit (SDK).

Keywords: LDWS, ADAS, Collision Avoidance, YOLO, SDK.