Stereo vision-based road obstacles detection

Abstract

This paper presents a fast road obstacle detection system based on stereo vision. The algorithm contains three main components: road detection, obstacle detection and obstacles tracking. The road detection is achieved by using a small rectangular shape at bottom center of disparity image to extract the disparities of the road. The roadsides are located by using morphological processing and Hough transform. In the obstacle detection process, the objects can be easily located by the segmentation process. The obstacles' tracking is achieved by the discrete Kalman filter. The proposed approach has been tested on different images. The provided results demonstrate the effectiveness of the proposed method.