Advances in Vision based Lane Detection Algorithm Based on Reliable Lane Markings

Abstract

This paper advances in a vision based driver assistance system for effective lane detection Technique. First method involves image processing technique initially to convert the Red, green and blue (RGB) into grayscale images takes place by removing the effects of shades and immersion data of the images by maintaining the brightness of the images. Then to remove the sound available in the images and the obtained video is allowed to pass through the Gaussian filter in-order to remove sound from the images. The edges recognition is an important function to identify and detect the sharp break in an image by canny edge detection algorithm. Then Hough transform technique is applied to obtain the image pixels in the form of long straight line in the Hough space. So we can detect the lane if the vehicle tends to deviate from the lane in a complex situation and take corrective measures to change the lane.