Road Surface Segmentation based on Vertically Local Disparity Histogram for Stereo Camera

Abstract

In this paper, we propose a road surface segmentation technique which is accurate and suitable for hardware implementation. The road surface is segmented by detecting the boundary between road and obstacle, based on the disparity histogram which we define as VLDH (Vertically Local Disparity Histogram). On each pixel of disparity image, VLDH is computed from the disparities of vertically local neighbourhood pixels. The major advantage is the feasibility of the pipeline processing on image processing hardware for stereo camera. The direct advantage on processing time is confirmed based on implementation into FPGA. Experimental result also shows that the accuracy of the proposed method is better than the conventional method.