Comparison of model-based polarimetric decomposition algorithms

Recently, many model-based decomposition algorithms, which are based on several scattering models, have been proposed to extract physical parameters of the Earth's surface [1]-[4]. Those algorithms decompose the received power into several scattering components, and some assumptions are necessary to find the scattering contribution. By the assumption(s) and scattering models in a used algorithm, decomposed scattering power of each component becomes often different with each other. Therefore, it is difficult to judge which algorithm is the best for a certain observation area. It can be useful to compare the algorithms and to clarify which one is suitable to detect a particular scattering mechanism. As a first step of the research, we compared several model-based polarimetric decomposition algorithms qualitatively from a viewpoint of “a visual difference in decomposition images”, “negative power”, and “ratio of scattering power”. In this paper, forested and urban area are selected for discussion.