Polarimetric SAR Image Classification Based on the Degree of Polarization and Co-polarized Phase-Difference Statistics

This paper proposes a polarimetric SAR image classification technique based on the degree of polarization (DoP) and co-polarized phase-difference (CPD) statistics. At first, the formulation for the DoP and CPD is derived. Then, the classification technique is verified with the JPL-AirSAR full polarimetric L-band data acquired in 2000, Korea. The technique has capability of classifying an SAR data into four major classes, such as bare surface, short vegetation canopy, tall vegetation canopy, and village.