Canny Detector and Hough Transform Based Layline and Region of Interest Detection for Autonomous Vechicles

Abstract

This paper introduces a lay line track detection algorithm for automated vehicles. Lay line detection of roads for automation vehicles is a simple computer vision technique in which the following steps are followed: processing a video, applying Canny Detector, segmenting lane area and finally Region of Interest (ROI) is produced from Hough Transformation. This work is carried out using OpenCV software. The methodology which was proposed in this paper, achieves sensitivity of 91.24% specificity, and 90.88% accuracy with a processing speed of 2ms.