Building Detection from Urban High-Resolution SAR Image Based on Facade Regularities

Inhalt:
The detection and extraction of buildings using high resolution synthetic aperture radar (SAR) images has been the topic of recent discussions. Given the structural details in the SAR image, algorithms for building extraction/monitoring should exploit this contextual information for an object-based analysis. In this paper, a framework for building detection based on point-like features is presented. a) A non-local strategy is integrated with a ratio edge detector to select point candidates; b) Based on strong scattering points, the Two-pass algorithm is used to generate line features. c) Low-level features are fused to identify individual buildings. The experimental results on our TerraSAR-X ST data set confirm that our scheme performs well.