Multi Class Image Segmenation Using Bottom-Up Approach
D. Kerana Hanirex · 2014
Multi-class image segmentation is one of the recent research areain image editing and content-based image retrieval. In this paper we are proposing a bottom up approach using JSEG, to segment an input image. It heuristically groups the pixels in the input image according to their spatial adjacency, feature proximity, boundary continuity, etc., However, as it focuses directly on the pixel or region it can obtain highly accurate region boundaries. But classical JSEG does not produce high quality segmentation for some class of images. Hence in this paper an approach of using I-FRAC and FRACTAL JSEG is proposed.