SAR calibration and principal component analysis

Abstract

Principal components analysis of ideal, complex quadpolarized synthetic aperture data tells us how the information is encoded in the eigenstructure of the data from homogeneous, distributed targets. Comparison with real data indicates distortion of the correlation matrices by system effects. The implications for measurement of the terms in the true correlation matrix are displayed. These are related to how accurately we can make inferences about the products of principle components analysis.