Vision-based Lane Departure Warning System

Abstract

Vision-based Lane Departure Warning System (LDW) is a promising tool to avoid road accidents. In practice, it is exceptionally hard to accurately and efficiently detect lanes due a variety of complex noise such as environmental variability. However, image processing techniques have shown promising and reliable outcome in detecting lanes during non- ideal conditions. Lane detection and lane departure measurement are two important modules in LDW system. This paper explores the gaps and limitations of the existing method in the past 10 years concerning lane detection and departure warning for LDW system.