Abstract: Aiming at the problems of low accuracy and slow convergence of traditional multilevel threshold image segmentation methods, an improved sand cat swarm optimization（ISCSO） algorithm was proposed for global optimization and applied to 2D-OTSU multi-threshold image segmentation task. By using Henon chaotic mapping and inverse refraction mechanism to initialize the population, the distribution of the population was made more uniform, and the starting state of the search was improved, so as to increase the global search capability of the algorithm; the nonlinear sensitivity update formula was adopted to balance the search diversity and convergence accuracy; the variable spiral search strategy was introduced to improve the position update algorithm, so as to ensure that the algorithm has better search diversity and the ability to jump out of the local optimal solution. The algorithm has the ability of searching diversity and jumping out of local optimal solutions.Six test images were selected for the 2D-OTSU multi-threshold image segmentation experiments of ISCSO algorithm, and the peak signal-to-noise ratio（PSNR）, feature similarity index（FSIM） and structural similarity index（SSIM） were used to evaluate the experimental results. And the experimental results show that, the result of 85. 2% obtained by using the 2D-OTSU multi-threshold image segmentation based on the ISCSO algorithm is better than the comparison algorithm. And the method has strong search accuracy and convergence speed.This proves the effectiveness and potential of ISCSO algorithm in the field of image segmentation.