| Research on automated road extraction has explored low and high resolution imagery as well as light detection and ranging (lidar) data. Many studies have considered fusion of high spatial resolution imagery and lidar data, while a small number of papers have reported road extraction from lidar data only. The goal of this study was to extract roads from lidar data and explore the impact of land use on the accuracy of the derived road network using six study sites in Oneida County, NY. The methodology developed considered two main steps: (1) defining raster road clusters, and (2) generating the vector road network. The quality assessment showed differences in road extraction accuracy for residential and commercial sites for both the raster and vector stages of analysis. The binary raster road delineation accuracy was better than that of the vectorization process, hence future research efforts will focus on improving the vectorization procedures.Key Words: Linear feature extraction, GIS updating... |
