K−way Hypergraph Partitioning and Color Image Segmentation

The goal of still color image segmentation is to divide the image into homogeneous regions. Object extraction, object recognition and object−based compression are typical applications that use still segmentation as a low−level image processing. In this paper, we present a method for color image segmentation. It formulates a color image segmentation problem as a partition of a Color Image Neighborhood Hypergraph (CINH) representation. Both global and local information are considered when we process this representation. To overcome the computational difficulty of directly solving the CINH partitioning problem, a multilevel hypergraph partitioning has been used. The proposed method is compared with the graph based segmentation algorithm using normalized cut criteria. The experimental results demonstrate the effectiveness of the proposed approach.