Fast InSAR Phase Unwrapping Method for Complex Mountainous Areas With High Noise and Large Gradient Changes

Abstract

Interferometry synthetic aperture radar (InSAR) is essential for monitoring surface deformation and geological hazards. InSAR applications in complex mountainous regions are challenging due to high noise levels and large phase gradient variations, which make it difficult to obtain accurate and efficient phase unwrapping results. As a result, the development of fast and high-precision phase unwrapping methods in these areas has become an urgent research topic in the current InSAR field. In response to this challenge, this study proposes a fast and high-precision phase unwrapping method specifically designed for complex mountainous regions. The method includes a sliding window singular value denoising phase gradient correction unwrapping algorithm (SDPC) for obtaining precise phase unwrapping results. A fast-phase unwrapping algorithm based on FPUNet networks is proposed by integrating deep learning techniques with the SDPC results. By integrating these two methods, the phase unwrapping speed can be accelerated while maintaining the unwrapping accuracy. The Dongchuan District of Kunming City, Yunnan Province, is used as the subject of this article. Because the true phase cannot be known using the measured interferometric phase data, simulated data are introduced to verify and demonstrate the validity and accuracy of the method presented in this article. Experimental results demonstrate the effectiveness of the proposed method for phase unwrapping, even in complex mountainous regions characterized by high noise and large gradient variations. Moreover, this approach significantly enhances the phase unwrapping speed. The presented method addresses the inefficiency and inaccuracy of InSAR in phase unwrapping in complex mountainous regions.

Keywords

• Complex mountainous area
• deep learning
• high noise
• interferometry synthetic aperture radar (InSAR) phase unwrapping
• large gradient changes