Improved spin filtering algorithm based on wavelet decomposition

Abstract

The invention discloses an improved spin filtering algorithm based on wavelet decomposition. The improved spin filter algorithm includes steps of performing decomposition by wavelet to obtain a low-frequency image carrying most of image information; spin-filtering the low-frequency image, defining the strip direction of the low-frequency image, obtaining the mean value of the direction lines, calculating the difference value of the grey values of the directions of a point and the mean value of the direction line, taking the absolute values, and accumulating to get a sum. The direction with the minimum accumulated sum and the tangent direction are found, and the median or mean value filtering can be carried out on the direction. According to the invention, the noise reduction effect is improved while the time efficiency is better than the original spin filtering, and furthermore a good prerequisite is provided for the phase extraction and phase unwrapping of interference image.

Description

technical field [0001] The invention relates to the technical field of computer-based electronic speckle interferometry, in particular to the noise reduction technology of speckle interference fringe patterns. Background technique [0002] In the initial research, speckle has been removed as noise that affects the image. Until 1966, Ennos found that it has measurable intensity and accurate phase, which determined the application value of speckle. Speckle is not only a kind of noise, but also an information carrier for measuring objects. [0003] People generally refer to electronic speckle interferometry (ESPI) processed by electronic hardware and digital speckle interferometry (DSPI) processed by digital software as electronic speckle interferometry. The basic principle is to let the laser speckle carry the information of the change of the measured object, and measure the phase change before and after the beam according to the relevant fringes of the speckle field generate...

AI Technical Summary

Problems solved by technology

The traditional filtering technology removes the noise part of the fringe image while reducing the contrast of the fringe image to make the image blurred, and sometimes destroys the structure of the fringe, which is not conducive to the next step of extracting the phase for measurement.

[0005] In the prior art, it is proposed to find similar pixels in the whole image instead of the image rotation filter in the local window, and the filtering effect is more effective, but this method takes a relatively long time to search in the whole image, and the algorithm speed is slow

[0006] In the prior art, the application of the partial differential equation method in speckle interference fringe image filtering is proposed, the image to be processed is regarded as a partial differential operator, and the initial image is deformed by using the partial differential equation, thereby linking the partial differential equation with the image , and achieved good results, however, this method requires more parameters and takes longer to process

[0007] At present, the research and analysis of speckle fringe image denoising at home and abroad does not have a method that can adapt to all speckle interference fringe images

Images