In order to improve the accuracy of insulator image segmentation and to resolve the problems of the traditional k-means algorithm that is sensitive to the cluster center, the number of clusters, and is prone to segmentation errors, an insulator image segmentation method based on improving the Hunter Pery Optimization-Gray Histogram (IHPO-GH) and optimizing the k-means algorithm was proposed. The gray histogram of the image is constructed, and the number of clustering centers of k-means algorithm is determined by obtaining the extreme points of the gray histogram; The IHPO algorithm based on the cosine decreasing method to balance the global and local search and the exponential decreasing method to improve the local solution accuracy is proposed to optimize the selection of the initial cluster center; through the above two methods, more accurate initial parameters can be obtained and the iterative process of k-means can be optimized. Experimental verification shows that the algorithm in this paper is significantly faster than the k-means algorithm in terms of iteration speed, and the solution accuracy is also greatly improved by calculating PNSR and MSE parameters. Finally, the visual effect analysis verifies the accuracy of the proposed algorithm image segmentation.