Signature-Based Target Detection for Dual and Compact Polarimetric SAR Applications

A signature-based target detection technique for partial and full polarimetric synthetic aperture radar (SAR) (PolSAR) applications is developed. This method is based on maximum-likelihood (ML) matching of multipixel polarimetric target signatures and provides optimal detection considering the multilook and multichannel nature of actual target’s SAR signature. It is shown that the signature-based detection can also aid target classification and recognition by the virtue of closest signature-class matching.