An improved lane and vehicle detection method in Driver Assistance System with Lane Departure and Forward Collision Warning

Abstract

The objective of this paper is to develop an advanced driver assistance system with lane departure warning and forward collision warning functions with enhanced method, which will speed up the computations. In order to extract lane markings and vehicles from roadway images captured by stereo camera, the image processing methods such as coordinate systems transformation, object detection and object tracking are applied to recognize the lane markings and the vehicles. In lane marking recognition, gray scale data, dynamically changing region of interest and featured based techniques are used to detect lane markings successfully. Experimental results show that the proposed algorithm is effective in image preprocessing and can detect the lane marking and vehicle accurately with less time.