Abstract: In order to deal with the problem that the traditional fuzzy cluster segmentation algorithms were extremely sensitive to noise and could not determine the cluster number automatically, the algorithm of fuzzy ISODATA image segmentation integrating Voronoi tessellation HMRF model is proposed. It divided the image domain into many sub-regions by Voronoi tessellation, and defined objective function with sub-regions based on Hidden Markov Random Field (HMRF) to reduce the effect of noise. Then, the cluster number was changed by Iterative Self-Organizing Data Analysis Techniques Algorithm (ISODATA) with cluster splitting and merging operations. Comparing the segmentation results of simulated, composite and real images from qualitative and quantitative analyses indicate that the proposed algorithm can not only overcome the effect of the image noises and outliers, but also obtain correct cluster number adaptively, and realize accurate image segmentation.