Fereshteh Massoumi - Islamic Azad university Zanjan Branch
Mahdi Afzali - Islamic Azad university Zanjan Branch

Roads are the most important part of transportation system. This importance led to many tries to extract road from satellite images and image mining. In this paper, we tried to extract roads in a dense urban area by using of image mining methods. Due to spectral similarity of urban objects in dense areas, there is no assurance to detect the urban objects properly based on spectral information. Therefore in current work, it is aimed to take advantage of two data sets including lidar data and aerial images. The proposed method increased accuracy and pace of road detection and also led to promote the overall accuracy till 95 percent

Image mining, Lidar data, Object base, Road detection