Research On Micro Displacement Measurement Base On Digital Image Cprrelation Method

As a high-precision non-contact full-field displacement and strain measurement method,digital image correlation method has been popularized and applied in aerospace,non-destructive testing,deformation measurement under high temperature environment and micro-displacement measurement.The principle of digital image correlation method is to search and calculate the surface speckle pattern before and after the deformation of the specimen.The factors affecting the solution efficiency and anti-interference are mainly: change of ambient illumination,Image quality,feature selection,selection of search algorithms and size of the calculation window.Classical methods have some shortcomings in solving efficiency and anti-interference.In order to make the digital image correlation method meet the engineering application requirements better,the following improvements are proposed in this paper.Aiming at the problem that the traditional digital image correlation method has great influence on the light change of environment,a digital image correlation method based on the central symmetric local binary mode(CS-LBP)is proposed.Firstly,the CS-LBP feature of the image template sub-region before deformation is extracted.According to the feature,the target sub-region with the highest matching degree is found in the deformed image,then the relative relationship between the template sub-region and the target sub-region in the image coordinate system is utilized.Calculate the displacement information of the object.The results of simulated speckle experiment and rigid body translation experiment show that the method effectively improves the robustness of digital image correlation method in light intensity variation.Aiming at the problem that the Newton-Raphson(N-R)method calculates the sub-pixel displacement time and affected by the initial value of displacement,a digital image correlation method based on SURF and GPU acceleration is proposed.First,according to the number of effective pixels of the image,the device side performs task division on the host side,then the device end uploads the calculated SURF feature to the host end.According to the SURF feature of the two speckle images before and after the deformation,the displacement value is matched.Finally,the displacement value is used as the initial value of the N-R iteration,finally,displacement field information is obtained.The cylinder compression and image acquisition verification experiments were carried out by using a vertical loading press.The results show that the proposed algorithm is moreefficient than classical algorithms.