Implementation of Lane Detection Algorithm for Self-driving Vehicles Using

Abstract

Recently, systems for detecting and tracking moving objects from video are gaining research interest in the field of image processing, owing to their applications in fields such as security, observation, and military, and considerable research is being conducted to develop high-accuracy and high-speed processing systems. In particular, as interest in autonomous driving has increased rapidly, various algorithms for lane keeping assistance devices have been developed. This study proposes a lane detection algorithm by comparing color-based lane detection algorithms and using a lane detection algorithm based on representative line extraction. Edge extraction and Gaussian filters are applied for lane detection and a Median filter is applied for image noise reduction. The detection accuracy is improved by extracting the region of interest for the image based on four pointers. Finally, a Hough transform is applied to improve the accuracy of straight line detection, and an algorithm to extract representative lines is applied to increase the detection rate in shadow regions and dark areas. Experimental results show that the proposed algorithm can detect lanes with high accuracy. The application of this algorithm to vehicle black boxes or autonomous driving will help prevent lane departure and reduce accident rates.