Lane Detection for Departure using Hough Transform for Driving System

Abstract

The Hough transform is technique of feature extraction used for analysis of images, computer vision, and digital image processing. The purpose of the technique is to find imperfect instances of objects within a certain class of shapes by a voting procedure. This voting procedure is carried out in a parameter space, from which object candidates are obtained as local maxima in a so-called accumulator space that is explicitly constructed by the algorithm for computing the Hough transform.

A Hough Transform Based Lane Detection for Driving System has been developed to aid a driver in the lane departure decision-making, to reduce a loss of concentration and to prevent an accident while driving. In this paper, we propose a method for detecting the lanes by using a webcam camera to record the road as a video file (.avi) and using Hough Transform method to detect the lanes in an image. The results show the lane detection in various line road conditions. According to the results, the drivers can use this information to increase their safety of driving, especially when making the decision of lane changing.