Abstract
As the number of traffic accidents is increasing in the world, collision prevention systems are available in the market. It is effective to know the road condition (wet or dry) for the system to enhance its performance. Recently on-board camera systems are becoming popular for vehicle safety systems and road surface detection methods that detect polarized light or reflecting infrared ray have been reported. In this paper, we propose a road surface detection method by applying Gabor filter to mono-camera images. We use normalized Gabor features in order to judge the road surface condition properly and showed that the accuracy of the proposed method is around 80% by testing in several different types of roads.
Keywords: Gabor filter, Detection of road surface condition