Vision based lane detection and departure warning system

Abstract

The increasing number of road accidents is a serious issue in front of modern society. Driver inattention, fatigue and drowsiness are the major causes of road accidents. Numerous methods have been proposed for lane detection and tracking. The main objective of this paper is comparison of three widely used lane detection techniques depending upon their similarities and differences, advantages and disadvantages. Hough transform, Road model and Fuzzy logic based techniques. Hough transform is most effective technique for detection of straight line with the advantage of having reduced logic area and memory utilization and high reliability. Model based techniques uses mathematical model for lane boundary fitting and requires very few parameters for lane representation because of this these systems are more robust against occlusion and missing data than feature based techniques. Fuzzy logic is very useful for drawing precise conclusions from imprecise input data and for solving problems which are difficult or impossible to model using exact mathematical model but can be easily solved by experience of human operator.