Fuzziness and Multi-neighborhood Strategy for Ridgeline Extraction DEMs Based

Abstract

Landform feature lines, as one kind of descriptors of landform skeletons, have been used in many fields, such as surveying, hydrology simulation, landform classification, and vegetation mapping. Due to the fuzziness and multi-scale characteristics of landform features, there are still some obstacles in the process of landform features exaction. The main weakness includes spatial extent, the vagueness of boundaries of landform feature lines. In this study, some efforts have been made to choose a relatively objective spatial extent in the ridgelines extraction. A multi-neighborhood method combined with fuzziness was used in the ridgelines extraction based on DEMs. The process included the following steps, choice of spatial scale, extraction watershed divides, determination degree of membership and extraction ridgelines according to given degree of membership. The result shows that this novel method can depict landform features in fine detail and the choice of spatial extent is more objective and reasonable.