Adaptive filter for unwrapping noisy phase image in phase-stepping interferometry

An adaptive filter for preprocessing wrapped phase image with high noise co ntent has been developed. The window size of this median filter is adjustab le according to its location in the noisy phase image. This size-adjustable median filter has not only high noise-rejection ability, but also both goo d phase-jump identification and image-detail preservation properties. Phase unwrapping can then be done to recover the real phase image without the ha mpering influence of noise, especially the noise near the phase jumps. Both numerical simulation and experimental results used to examine the effectiv eness of this adaptive filter are presented. (C) 2001 Published by Elsevier Science Ltd.