Application of image processing to detect lane for autonomous vehicle

Abstract

Image processing has many applications in intelligent transportation system. Lane detection and tracking for autonomous vehicle is the one of that. This study uses Matlab software and it’s Toolbox: Image Processing, Image Acquistion System and Computer Vision System as the main tool to collect and process. The road model is assumped that: texture of road are identical. Lane makers follow lane rules. The distance between the lane makers is constant. This project uses webcam to collect the images. From the images obtained by using the 2D FIR filter, grayscale image is achieved. The image are converted into binary images, thenn by using Canny method and Hough transform available in Matlab to determine the lane makers and lane deparment. Calculate the distance based on separator obtained to provide signal for autonomous vehicle via the standard RS232 interface. The experimental results on local streets show that the sug- gested program is very reliable