The research of road extraction for high resolution satellite image

High-resolution satellite remote sensing image plays a significant role in Geographic Information System (GIS). Future GIS applications require high spatial resolution image, which cannot be met by traditional satellite image with resolution of 20 m or lower. In this paper, integrating high-resolution SPOT-5 data into GIS is discussed. Various methods have been used to extract roads from satellite images with different resolutions. In this paper, after a road is extracted from a SPOT-5 satellite image, the extracted roads are converted into road vectors in order to use these vector road maps in GIS. We propose a three-step method for road extraction. The first step is to automatically detect the road edge of the interested region. As the second step, we transform the road edge into vectors. In the third step, the transformed road vectors are compared with the existing official road vectors, and then some inaccuracies will be corrected. Through these steps, we can acquire accurate road vectors. Using extracted road vectors, which can be used to rapidly complement new roads for existing maps and to provide integrated road information of a city, we can conveniently provide timely location service for telecommunication.

[1]  Donald Geman,et al.  An Active Testing Model for Tracking Roads in Satellite Images , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  Taejung Kim,et al.  Semi-Automatic Road Extraction Algorithm from IKONOS Images Using Template Matching , 2001 .

[3]  D Marr,et al.  Theory of edge detection , 1979, Proceedings of the Royal Society of London. Series B. Biological Sciences.

[4]  David B. Cooper,et al.  Automatic Finding of Main Roads in Aerial Images by Using Geometric-Stochastic Models and Estimation , 1996, IEEE Trans. Pattern Anal. Mach. Intell..