UnwrapDiff: Conditional Diffusion for Robust InSAR Phase Unwrapping

2025-12-04

AI Summary

InSAR phase unwrapping is severely degraded by radar noise and decorrelation, limiting the accuracy of deformation monitoring. To address this, we propose the first conditional diffusion model (CDM) specifically designed for InSAR unwrapping, which leverages SNAPHU’s output as a physics-guided prior and synergistically integrates denoising diffusion probabilistic models (DDPM) with a minimum-cost flow framework—thereby preserving physical consistency while significantly enhancing noise robustness. We further introduce a novel synthetic dataset incorporating realistic atmospheric disturbances and diverse noise types for rigorous training and evaluation. Quantitative experiments demonstrate that our method achieves a 10.11% average reduction in normalized root-mean-square error (NRMSE) over SNAPHU on synthetic benchmarks. Notably, it exhibits superior robustness and reconstruction fidelity in challenging scenarios characterized by strong noise and high-gradient deformations—such as volcanic uplift—where conventional methods often fail.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Intelligent Robots: State Estimation
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Phase unwrapping is a fundamental problem in InSAR data processing, supporting geophysical applications such as deformation monitoring and hazard assessment. Its reliability is limited by noise and decorrelation in radar acquisitions, which makes accurate reconstruction of the deformation signal challenging. We propose a denoising diffusion probabilistic model (DDPM)-based framework for InSAR phase unwrapping, UnwrapDiff, in which the output of the traditional minimum cost flow algorithm (SNAPHU) is incorporated as conditional guidance. To evaluate robustness, we construct a synthetic dataset that incorporates atmospheric effects and diverse noise patterns, representative of realistic InSAR observations. Experiments show that the proposed model leverages the conditional prior while reducing the effect of diverse noise patterns, achieving on average a 10.11% reduction in NRMSE compared to SNAPHU. It also achieves better reconstruction quality in difficult cases such as dyke intrusions.

Problem

• Robust InSAR phase unwrapping under noise and decorrelation
• Improving deformation signal reconstruction from noisy radar data
• Enhancing reliability of geophysical monitoring and hazard assessment

Innovation

• Denoising diffusion model for InSAR phase unwrapping
• Conditional guidance using traditional algorithm output
• Synthetic dataset with realistic noise for evaluation