Contrast Measures based on the Complex Correlation Coefficient for PolSAR Imagery

TL;DR

This paper introduces new contrast measures for PolSAR imagery based on the complex correlation coefficient, providing statistical tools for hypothesis testing and validating them with real data.

Contribution

It develops novel contrast measures involving the number of looks and complex correlation, along with confidence regions and hypothesis tests for PolSAR data.

Findings

Contrast measures effectively differentiate polarization channels.

Statistical confidence regions enable hypothesis testing.

Application confirms practical usefulness of the proposed measures.

Abstract

We derive contrast measures which involve the number of looks and the complex correlation coefficient between polarization channels in PolSAR imagery. Using asymptotic results which characterize the behavior of these measures, we derive statistical regions of confidence which lead to test of hypothesis. An application to real data is performed, confirming the importance of the proposals.