Different approaches to multiedge detection in SAR images

Edge detection is a fundamental issue in image analysis. The presence of speckle, which can be modelled as a strong multiplicative noise, complicates edge detection in synthetic aperture radar (SAR) images. Several statistical edge detectors have been developed specifically for such images, based on the hypothesis that only one step edge is present in the analyzing window. The authors concentrate on the spatial aspect of edge detection in SAR images. Depending on the scene type and the size of the analyzing window, the monoedge hypothesis is violated more or less frequently. They present and compare three approaches to edge detection in a multiedge context: an optimum non-adaptive operator, a spatially adaptive operator and multiscale analysis