Branch-cut algorithm with fast search ability for the shortest branch-cuts based on modified GA

Abstract

The branch-cut algorithm is a powerful noise-immune algorithm for two-dimensional phase unwrapping. Shorter branch-cuts lead to better results for the unwrapped phase. Herein, we propose a branch-cut algorithm with fast search ability for the shortest branch-cuts based on modified GA. First, the local-nearest-neighbor algorithm is used to pair the positive and negative residues so as to optimize the initial population. This step is fast but gives a local optimum. Next, a branch-cut algorithm based on modified GA is used to globally search the paired residues giving branch-cuts with the shortest total length. Finally, the phase is unwrapped while avoiding branch-cuts. The performance of the proposed algorithm is tested by both simulation and experiment, and the results demonstrate that it can rapidly find the branch-cuts with shortest total length and a high solution accuracy.

KEYWORDS:

• Two-dimensional phase unwrapping
• local-nearest-neighbor algorithm
• improved branch-cut algorithm

Additional information

Funding

This work was funded by the National Natural Science Foundation of China (51275405) and the Natural Science Foundation of Shaanxi Province (2015JM6333).