Abstract
In this study, we assessed the feasibility of road area extraction from LIDAR height and intensity data. We tested two methods. One is a maximum likelihood method using color image and LIDAR data. The other is the method using region segmentation. At first, a height based region segmentation method was applied and flat regions were extracted. Then, each region was distinguished by average intensity in the region and the road candidate regions were picked up. The results of two methods were compared with the result of a maximum likelihood method using color image. It was revealed that the extraction result of the proposed methods was better than that of the reference method. From this study, we concluded that the road extraction with LIDAR height and intensity data could be practicable, though the vectorisation process remains to be solved.
