Phase unwrapping based on a parallel noise-immune algorithm

摘要

A newly developed path-independent phase unwrapping algorithm, which is simple and concise, is presented. The algorithm is based on a parallel elimination of local phase jumps and the subsequent shifts of them, through iterations, to the boundary or to cancel with others in the midway. The number of iterations never exceeds the sum of the width and height of the processed map, whereas for cellular automata algorithm, a multiplier of the cycle of iterations by the proposed algorithm herein is easily needed for successful treatment of a same map. The proposed noise-immune algorithm is applied to a phase map of electronic speckle pattern interferometry (without any filtering work) to prove its effectiveness. (C) 2002 Elsevier Science Ltd. All rights reserved. [References: 26]