The Implication of Boundary Adjustment in H/Alpha Plane on the Unsupervised Wishart Maximum Likelihood Segmentation for Fully Polarimetric SAR Data

In the practical work of the H/Alpha Wishart Maximum Likelihood segmentation, Cloude-Pottier's original boundary scheme would have often been a default setting. In this paper, the implication of boundary adjustment in H/Alpha plane on the unsupervised Wishart ML segmentation for fully polarimetric SAR data is researched. Four boundary adjustment schemes are designed, and the Wishart ML segmentation results are compared with that of the original boundary scheme. The analysis illustrates that boundary setting in H/Alpha plane is an important factor affecting the H/Alpha Wishart ML segmentation performance.