Unsupervised classification for hybrid polarimetric SAR data based on scattering mechanisms and Wishart classifier

An unsupervised classification algorithm utilising both polarimetric scattering mechanisms (PSMs) of hybrid-polarity data and the Wishart classifier is proposed. The initial scattering categories of the proposed algorithm are derived from the roll-invariant m-χ classification algorithm. Pixels with no clearly defined dominant PSM are excluded, and the resulting categories are expanded into a specified number of classes. These derived classes are taken as training samples of the Wishart classifier. The effectiveness of the proposed algorithm is validated with the dataset over San Francisco.