Robust phase unwrapping via non-local regularization.

Area of Science:

• Optics and Photonics
• Image Processing
• Computational Imaging

Background:

• Phase unwrapping is crucial for reconstructing true phase information from wrapped phase data.
• Conventional methods struggle with severe noise, leading to error propagation and reduced accuracy.
• Robust phase unwrapping is essential for various imaging applications.

Purpose of the Study:

• To develop a robust and high-fidelity iterative phase unwrapping framework.
• To address the limitations of conventional methods in noisy environments.
• To enhance the accuracy and efficiency of phase recovery.

Main Methods:

• Utilizes the transport-of-intensity equation for efficient phase unwrapping.
• Incorporates non-local structural similarity via low-rank regularization for improved accuracy.
• Employs an adaptive iteration strategy with dynamic denoising parameter updates to preserve image details.

Main Results:

• The proposed method demonstrates high fidelity and robustness under severe noise conditions.
• Achieves significantly improved reconstruction accuracy compared to existing methods.
• Outperforms state-of-the-art techniques, showing at least 6 dB higher peak signal-to-noise ratio (PSNR).

Conclusions:

• The novel iterative framework offers a superior solution for phase unwrapping in noisy conditions.
• The combination of transport-of-intensity, low-rank regularization, and adaptive denoising enhances phase recovery.
• This method provides satisfying results and advances the field of phase retrieval in imaging.