Principal Component Analysis for polarimetric radar cross-section imaging

The decomposition of polarimetric radar cross-section images into Huynen parameters is useful for distinguishing different scattering mechanisms which can occur at a radar target. However, if several of these mechanisms superimpose within one resolution cell the result of this decomposition can be misleading. In this paper the combination of a decomposition into Huynen parameters with Principal Component Analysis is proposed. This combination allows to separate three decorrelated scattering centers to reveal scattering mechanisms covered by mechanisms with higher radar cross-section. The application to measurement data demonstrates the success of this approach.