Core image segmentation method by combining watershed and ISODATA algorithm

Abstract

Watershed based image segmentation can make the edge close and precise, but has the drawback of over-segmentation. ISODATA can conglomerate gravel, but has the edge bias error. Combining watersheds and ISODATA, this paper proposed an image segmentation method to solve these problems. Firstly, watershed segmentation algorithm was applied to get the initial over-segmentation image, and Iterative Self-Organizing Data Analysis Technique Algorithm (ISODATA) was applied to get the conglomeration image. Then, labeled images and inverse wavelet transform were used together to get the full-resolution image. The results based on ISODATA algorithm were applied to solve over-segmentation problem caused by watershed. Experimental results show that it is a practicable method for the image segmentation.