Supervised classification of coherent and incoherent polarimetric SAR observables: comparison and accuracy assessments

摘要

Classification of Earth surface using SAR observations constitutes an important application of polarimetric SAR. The most promising and investigated polarimetric parameters for such a task have been the Entropy, a and Anisotropy (H, a and A) parameters. A similar use of coherent methods, however, appears to have been scarcely considered and remained essentially untested. In this contribution, we wish to address this issue, testing and comparing a wide range of polarimetric SAR parameters, coherent and incoherent, by means of different classification algorithms. An original aspect of this work is also the study of the dependence of the classification results on the varying size of averaging windows of pixels. Such an analysis will permit to prove if the chosen polarimetric parameters provide a description only of "point-like" physical properties of the targets or if they also contain "extended", local information. The investigations to be reported should, therefore, not only reveal a systematic and quantitative assessment of the classification efficacy of different methodologies but also afford their comparison, an exercise, hitherto unavailable in the literature in common knowledge.