Interferometric Phase Improvement Based on Polarimetric Data Fusion

Area of Science:

• Remote Sensing
• Signal Processing
• Electromagnetics

Background:

• Interferometric phase quality is crucial for applications like terrain mapping and deformation monitoring.
• Polarimetric interferometric data often suffers from noise and reduced reliability due to low amplitude signals.
• Existing methods may not fully optimize phase quality across all polarimetric channels.

Purpose of the Study:

• To develop a novel method for improving interferometric phase quality in polarimetric SAR data.
• To enhance the reliability and accuracy of phase information by maximizing amplitude through data fusion.
• To provide a robust preprocessing step for phase unwrapping algorithms.

Main Methods:

• A pixel-wise fusion approach is proposed, projecting coherent polarimetric scattering vectors onto an optimal direction.
• The fusion strategy maximizes the minimum amplitude of the projected scattering vectors.
• In the single-look case, the fused phase is a weighted average of individual channel phases, offering physical interpretability.

Main Results:

• Significant reduction in phase noise and residue points without additional filtering.
• Demonstrated substantial improvement in interferometric phase quality compared to standard methods.
• Effectiveness and robustness validated using SIR-C/X-SAR data, outperforming Cloude's coherence optimization.

Conclusions:

• The proposed fusion method effectively enhances polarimetric interferometric phase quality.
• This technique serves as a valuable preprocessing tool, particularly before phase unwrapping.
• The approach offers a robust and physically meaningful way to improve SAR interferometry data.