[A novel pol SAR image classification method with subsequent category adjustment by terrain scattering...

Area of Science:

• Remote Sensing
• Geospatial Analysis
• Signal Processing

Context:

• Polarimetric Synthetic Aperture Radar (PolSAR) data provides rich information for land cover classification.
• Traditional supervised classification methods can be limited by spatial distribution and feature vector analysis.
• Accurate land cover classification is crucial for environmental monitoring and resource management.

Purpose:

• To propose a novel supervised classification method for PolSAR images.
• To enhance classification accuracy by integrating coherence matrix information and iterative correction.
• To improve the analysis of terrain scattering characteristics for better land cover categorization.

Summary:

• A supervised classification method is proposed, combining complex Wishart classifier ideology with coherence matrices for PolSAR image analysis.
• Initial classification using Cloude & Pottier polarimetric characteristics is refined through iterative correction with Kernel Fuzzy C-Means, guided by scattering similarity analysis.
• The method utilizes X-band full polarization SAR data and demonstrates improved classification accuracy and better representation of surface object scattering characteristics.

Impact:

• The proposed method achieves favorable classification accuracy for PolSAR images.
• It offers a more robust approach to land cover classification by considering scattering characteristics.
• This advancement can lead to more precise geospatial analysis and environmental monitoring applications.