A New Use of the Two-Look Internal Hermitian Product for Discrimination Between Natural and Artificial Objects

Inhalt:
Discriminating between natural and artificial objects is an important part of target detection and recognition. As synthetic aperture radar (SAR) images present features that complicate the task of image analysis, automated algorithms are of help. This paper describes an algorithm that makes use of both the radiometry and coherence information of SLC images. The results on both airborne and satellite images are promising, especially on high resolution images.