Non-contact Measurement Technique Based On Digital Image Correlation Method With Its Applications To Geotechnical Deformation

The Digital Image Correlation(DIC)method is a new non-contact full-field displacement deformation measurement optical method.Compared with other measurement methods,it not only requires lower measurement conditions but also has simple data acquisition and non-contact,High measurement accuracy,full-field measurement and other advantages.In recent years,with the advancement of computer performance and the improvement of camera pixels,DIC has been widely used in medicine,materials science,aerospace and other fields,and has good experimental results,and has made considerable development and progress.This paper summarizes the existing methods of measuring displacement and deformation,and compares their advantages and disadvantages in terms of measurement accuracy and ease of operation.The Digital Image Correlation method is the trend of development of displacement and deformation measurement,but there are still some defects such as low influence of ambient light changes and low measurement accuracy of test specimens that undergo rotational deformation.Based on the characteristics of geotechnical engineering tests,several key steps of DIC method have been improved and optimized to improve their measurement accuracy and calculation efficiency in geotechnical engineering tests.First of all,this paper proposes a grayscale aberration algorithm to reduce the influence of ambient light changes on the measurement results.It is found that The method significantly improves the accuracy and stability of displacement calculation when the light intensity changes within 40% and between 60%-100% through experimental comparisons.It introduces a displacement conversion calibration method for converting pixel virtual displacement to real displacement,and proposes a displacement conversion calibration method based on Hough transform line detection with obvious lens imaging distortion.The real displacement field in the experiment is obtained by multiplying the pixel equivalent field generated by it and the pixel displacement field.Secondly,by comparing the search time in point-level traversal search,cross search,and the integer pixel search method based on particle swarm optimization algorithm in images of various levels of resolution,this paper proposes a combined integer pixel search based on calculating the size of the image area.This method not only maximizes the speed advantage of the particle swarm optimization search method and the cross search method,but also avoids the errors caused by their trapping in local optimization to a certain extent.For higher resolution pictures or interpolation processing.The search speed of the integer pixels of the super-resolution pictures has been significantly improved.Third,in view of the discontinuity of geotechnical engineering materials and the deformation characteristics that are prone to strain localization,a rotation matching search process based on bilinear subpixel interpolation is added between the whole pixel search and the subpixel search process.Finally,the main program and graphical interactive interface were written based on PYTHON and three groups of tests were designed to verify the feasibility of the program.The sand displacement test and rigid body rotation test under unstressed state verified the program's application in the geotechnical test environment.The natural speckle digital image formed by the sand particles is well recognized.The simulation analysis of the displacement field measurement under the concentrated load of the sand foundation model verifies that the program in this paper can meet the measurement accuracy requirements under general circumstances.