Performance comparison for lane detection and tracking with two different techniques

Abstract

This paper presents the performance comparison for lane detection and tracking using two techniques; Hough transform and color thresholding. Hough transform includes the following steps: gray scale conversion, edge detection, Hough space accumulation and un-Hough step to draw line. Color thresholding algorithm comprises color thresholding, morphology filtering and drawing tracking label. The results showed that Hough transform technique outperforms color technique in terms of reliability, efficiency and speed of code implementation. Both techniques are coded using MATLAB R2010b.