Polarimetric Decomposition and Machine Learning‐Based Classification of L‐ and S‐Band Airborne SAR (LS‐ASAR) Data

The polarimetric Synthetic Aperture Radar (SAR) data sets have been widely exploited for land use cover (LULC) classification due to their sensitivity the structural and dielectric properties of imaging target. In this study, potential fully L- S-band Airborne SAR (LS-ASAR) were explored machine-learning-based Urban, Vegetation, Waterbody, Open Ground. This work was done by utilizing dual-frequency airborne Santa Barbara, California, USA, acquired under (LS ASAR) campaign, a precursor mission space-borne NASA-ISRO (NISAR) mission. LS-ASAR information utilized LULC using SVM classifier. roll-invariant Barnes, eigenvalue/eigenvector-based Cloude H/A/Alpha decomposition implemented retrieve scattering parameters. backscatter response classes studied, separability analysis reduce misclassification error between six class pairs- Vegetation—Urban, Vegetation—Waterbody, Vegetation—Open Ground, Urban—Waterbody, Urban—Open Water—Open models failed achieve desirable index all pairs; consequently, Cloude, showed vegetation-urban class, waterbody-open ground both sets. effort made toward improving accuracy integrating eigenvalue-eigenvector parameters multifrequency set. method presented class-pair; eventually highest achieved i.e. 93.35% ( = 0.91) significantly reducing pairs.