InSAR Phase Unwrapping: A Comprehensive Review of Algorithms, Challenges, and Future Directions
Abstract
Context: Phase unwrapping is a fundamental step in interferometric Synthetic Aperture Radar (InSAR), requiring the reconstruction of a continuous phase field from observations available only modulo-2π. Objective: This review critically analyzes the main methodological approaches for InSAR phase unwrapping and highlights their strengths, limitations, and research trends. Method: Existing methods are categorized into four families: classical single-interferogram, optimization-based, multi-baseline and network-based, and learning-based approaches. Their principles and performance are compared under challenging conditions such as noise, decorrelation, and phase discontinuities. Results: Classical methods are computationally efficient but sensitive to local errors. Optimization-based approaches improve global consistency at higher computational cost. Multi-baseline and network-based methods enhance robustness through redundancy, whereas learning-based methods provide flexibility but remain limited by generalization and consistency. Conclusion: The review identifies key trade-offs among existing methods and emphasizes the need for hybrid and adaptive frameworks to achieve robust, scalable, and reliable InSAR phase unwrapping.
