Research On In-plane Displacement Measurement Based On Digital Speckle Correlation Method

Digital speckle correlation method,also known as digital image correlation method,is a new optical measurement method based on computer vision and digital image processing technology.The basic principle is that through extracting the corresponding information of speckle images of original and deformed object surface to obtain displacement of the whole field.because of its advantages such as non-contact,simple path of rays,high precision full-field analysis,and so on.Since its appearance in the 1980s,it has drawn wide attention from domestic and foreign scholars At present,it has formed a relatively complete theoretical system,and has been successfully applied in many fields,such as:civil engineering,materials testing,aerospace,and so on.Thanks to the limitation of the algorithm itself,the digital speckle correlation method still exist many deficiencies in terms of displacement measurement accuracy and calculation speed.In order to get higher computational accuracy,we usually have to sacrifice the computational efficiency.In view of this situation,based on the existing research results,the dissertation carries out an in-depth research on the displacement measurement algorithm of two-dimensional digital speckle correlation method.The main contents are as follows:(1)The basic principle of digital speckle correlation method is expounded,including speckle image,measurement system,displacement calculation and so on.Two kinds of related algorithms are introduced:one is the correlation matching algorithm based on the grayscale of the image(grayscale correlation method),and the other is the matching algorithm based on the fractal features of the structure(the fractal correlation method).The correlation matching function is selected by using simulated speckle images.(2)A variety of the whole pixel and sub-pixel displacement measurement algorithms of digital speckle correlation algorithm are studied.Several integer pixel displacement measurement algorithms are analyzed,and the point-by-point search method is used to compare the coefficient surfaces of the grayscale correlation method and the fractal correlation method.Based on the MATLAB platform,a series of mathematical models of sub-pixel displacement measurement algorithms are established,including grayscale interpolation,surface fitting(based on grayscale correlation fitting and fractal correlation fitting)and other algorithms(gradient Method and N-R method).The search capabilities of the above-mentioned various sub-pixel algorithms are compared,and a sub-pixel algorithm of surface fitting method is proposed,which combines the fractal correlation method and the grayscale correlation method.(3)A set of digital speckle correlation calculation software which can carry out artificial interactive operation is written.The feasibility and accuracy of the improved algorithm are verified by computer simulation and dynamic deflection test.