Research of terrain classification using disparity value for UGV

In this research, began to research for create local map for the unmanned ground vehicle to create efficient driving path. Detected space that car can drive freely and to create local map based on information of the free-driving space. Generate disparity image using stereo image and by utilizing the uv-disparity domain that has been converted from disparity image to classify surrounding environment, obstacles space and ground space. Finally, create local map by using the free-driving space that classified so can it be referenced when generate driving path.