The extended digital image correlation based on intensity change model

•A set of speckle image matching algorithms that add intensity model is proposed.•Linear least squares estimation is used for integer pixel matching algorithm.•Sub-pixel matching algorithm is derived combined IC-GN with intensity model.•It is suitable for illumination variation and can improve measurement accuracy. Digital image correlation is used to measure deformation and shape. In some complex scenes, the illumination of the measured surface may vary unevenly before and after deformation. The intensity change model is added to correlation function and a family of speckle image matching algorithms is obtained. In integer pixel matching, linear least square estimation is used to calculate the initial values of the intensity coefficients and integer pixel position; in subpixel matching, cyclic variable method is proposed to make the iteration converge. Compared with the current mainstream algorithm, the proposed method makes the physical meaning of correlation matching clearer, so that the correlation function can be defined according to the actual lighting environment, and correlation matching can be more accurate. The experimental results indicate that the proposed matching algorithm is suitable for applications where ambient light is constantly changing.