Frequency domain filtering for phase fringe patterns of digital speckle pattern interferometry

Abstract

In order to solve the problem of serious influence of much noise in the phase fringe patterns of digital speckle pattern interferometry on the result and accuracy of phase unwrapping,an effective method of stripe sine / cosine decomposition combined with frequency domain low-pass filtering is introduced in this paper,and the high precision filter of phase fringe patterns of digital speckle pattern interferometry is realized. Basic idea of this method is that phase fringe patterns is mapped into two patterns by the sine / cosine function before the phase image filter processing,and then is synthesized into phase fringe patterns through frequency domain filter. This decomposition frequency domain filtering method can retention phase jump information effectively when filtering. Experimental results show that compared with the traditional image noise reduction method,this method can filter out the speckle noise in images better on the basis of reserving "rush"image information,and it is simple and effective,which solves the problem of loss from 10% to 40% for phase fringe patterns gray level information when traditional filtering methods are used in stripe phase fringe patterns.