Unsupervised classification of PolSAR images using eigenvector analysis, Krogager decomposition and the Wishart classifier

In this paper, a new scheme for unsupervised classification of polarimetric synthetic aperture radar (PolSAR) images is presented. The method mainly consists of four parts: eigenvector analysis of the coherency (or covariance) matrix, Krogager decomposition, unsupervised classification using Krogager coefficients and scattering entropy, and iterative classification based on the Wishart distance measure. The method can classify the pixels into nine classes, and its effectiveness is demonstrated using the Jet Propulsion Laboratorypsilas AIRSAR and SIR-C/X-SAR L-band PolSAR data.