Comparison of Methods for Target Detection and Applications Using Polarimetric SAR Image

摘要

Polarimetric SAR (PolSAR) is sensitive to the orientation and characters of object and polarimetry could yield several new descriptive radar target detection parameters and lead to the improvement of radar detection algorithms. Target decomposition theory has been used for information extraction in PolSAR, and it can also explore the phase message in PolSAR data. In this paper, a comparison of polarimetric target decomposition methods is proposed. We generate a validity test for these methods using DLR ESAR L-band full polarized data. Results show that among many target decomposition algorithms, the coherent and incoherent formulations are quite comparable in distinguishing natural targets and man-made buildings. Pauli decomposition, Cameron decomposition and Freeman decomposition are suitable for the detection of natural targets. On the other hand, SDH decomposition, OEC decomposition, and Four-component model, in particular, are very useful for man-made target extraction.