Research On Digital Image Correlation Method Based On Hermite Sub-Pixel Gray Reconstruction

After years of development,digital image correlation methods have been widely used in many research and application fields.In the digital image correlation measurement technology,the sub-pixel displacement measurement algorithm of the digital image correlation method has become the research focus of the field because the whole pixel displacement measurement cannot meet the measurement accuracy requirement.In the digital image correlation sub-pixel displacement measurement algorithm,Newton-Raphson iterative algorithm(N-R iterative algorithm for short)has received extensive attention because of its high computational accuracy and good computational stability.In the sub-pixel displacement calculation process of the traditional N-R iterative algorithm,the cubic spline interpolation calculation method is usually used to reconstruct the gray value of the sub-pixel position.In the interpolation algorithm,the cubic Hermite interpolation algorithm has a good smoothness.Therefore,in order to further improve the accuracy of sub-pixel displacement solution of N-R iterative algorithm,this paper proposes an N-R iterative algorithm based on Hermite sub-pixel gray-scale reconstruction and studies its related characteristics.(1)Based on the accuracy of sub-pixel grayscale reconstruction of traditional N-R iterative algorithm,an N-R iterative algorithm based on Hermite sub-pixel grayscale reconstruction is proposed.The algorithm uses the Hermite interpolation function to establish the sub-pixel gray reconstruction method,and the Barron operator is used to calculate the first-order gray gradient.Finally,the Newton iterative algorithm is used to solve the sub-pixel displacement.The feasibility of the proposed algorithm is verified by numerical deformation experiments of real speckle patterns.Compared with the traditional N-R iterative algorithm,the proposed algorithm has better computational accuracy.(2)In order to further improve the computational efficiency of the N-R iterative algorithm for Hermite sub-pixel grayscale reconstruction,and avoid the repeated calculation of the first-order grayscale gradient in the sub-pixel displacement measurement calculation,this paper proposes a global grayscale gradient calculation strategy.At the same time,in order to improve the initial value precision of Hermite sub-pixel gray-scale reconstruction NR iterative algorithm,the influence of initial value on the calculation accuracy and efficiency of sub-pixel displacement solution is avoided.According to theprinciple of surface deformation continuity,the seed-based point is proposed.The search strategy,which uses the sub-pixel position of the point to be the initial value of the sub-pixel displacement of the adjacent point to be sought.The effectiveness of the two strategies is verified by numerical simulation experiments.The application of the two strategies effectively improves the computational efficiency of the N-R iterative algorithm without changing the computational accuracy of the algorithm.(3)The error analysis of the N-R iterative algorithm based on Hermite sub-pixel gray reconstruction is carried out.The function of sub-pixel position and bit value error is established by the formula of bit value error.At the same time,the simulation experiment is carried out on how the size of the calculation window affects the calculation accuracy of the N-R iterative algorithm proposed in this paper.The results show that the mean error of the NR iterative algorithm in different calculation windows is different.At the same time,the mean error of each speckle pattern in the same calculation window is different,which indicates that the difference of the calculation window size will be proposed in this paper.The NR iterative algorithm produces an impact.The result of the maximum absolute error can be obtained.The calculation window is more suitable for the N-R iterative algorithm proposed in this paper when it is in the pixel range.Finally,three commonly occurring random errors are analyzed and suggestions on how to reduce the effects of random errors are given.(4)On the mechanical performance test system based on dual camera tracking,the N-R iterative algorithm of Hermite sub-pixel gray reconstruction was used to analyze the tensile test of the material.The experimental results show that the proposed NR iterative algorithm has less error compared with the results of the extensometer measurement,and the proposed algorithm has higher accuracy,which can meet the measurement requirements,and verifies the feasibility of the proposed NR iterative algorithm and practicality.