A Thinning-Based Method for Extracting Peri-Urban Road Network from Panchromatic Images
The extraction of linear elements from remotely sensed imagery is a quite complex process. The extraction, for example, of the road network requires also its recognition which is not an easy operation, because road segments are not represented by a constant range of digital numbers in the input images, the shapes of the roads are illdefined, and pixels with the same digital number often represent land cover types which are not related to roads. That happens because in remotely sensed images, certain types of land cover, e.g. buildings, have similar spectral characteristics with roads and, thus, are indistinguishable, at a pre-processing stage, from the roads.
KeywordsRoad Network Binary Image Land Cover Type Road Segment Thresholding Algorithm
Unable to display preview. Download preview PDF.
- 3.Schenk T., and Ofer Zilderstein, 1990, Experiments with a Rule-Based System for Interpreting Linear Map Features. Photogrammetric Engineering and Remote Sensing, 56, 911–917.Google Scholar
- 4.Cleynenbreugel V.J. Fierens F. Suetens P. and Oosterlinck A. 1988 Knowledge-Based Road Network Extraction on SPOT Satellite Images. Proceedings of the Fourth International Conference on Pattern Recognition Cambridge U.K. March 1988 (New York: Lecture notes in Computer Science No301 Springer Verlag) pp. 352–359Google Scholar
- 6.Karathanassi V., Iossifidis C., Rokos D., 1998, A Thinning-Based Method for Recognizing and Extracting Peri-Urhan Road Network from SPOT Panchromatic Images. International Journal of Remote Sensing. (to be published)Google Scholar