Fine-Resolution Polsar Data Analysis with Compensation Theory for Improved Interpretation
Main Article Content
Abstract
Polarimetric Synthetic Aperture Radar (PolSAR) imaging is a powerful tool for analysing mechanisms of surface scattering, but its accuracy is often compromised by Polarization Orientation Angle (POA) shift. The POA shift occurs due to terrain slopes, vegetation structures, built-up and other structural features, alters scattering mechanisms, and affects decomposition-based analysis. This study used the Yamaguchi four-component decomposition model to investigate the impact of POA shift on EOS-04 (RISAT-1A) Fine Quad-Polarization data, acquired over Nashik District, Maharashtra, India. The investigation was performed before and after POA correction to analyze changes in even-bounce, odd-bounce, and volume scattering components. The statistical and histogram results show that without POA correction, volume scattering was exaggerated while double-bounce scattering was underestimated, resulting in inaccuracies in land cover classification and urban mapping. POA correction enhanced classification accuracy and enabled more reliable identification of scattering parameters. This demonstrates the potential of EOS-04 dataset and also importance of applying POA correction to PolSAR data during the preprocessing stage to obtain more accurate decomposition results, more precise geophysical parameters, and reliable land cover classification.