Normalized cuts and image segmentation

We propose a novel approach for solving the perceptual grouping problem in vision. Rather than focusing on local features and their consistencies in t he image data, our approach aims at extracting the global impression of an image. We treat image segmentation as a graph partitioning problem and prop ose a novel global criterion, the normalized cut, for segmenting the graph. The normalized cut criterion measures both the total dissimilarity between the different groups as well as the total similarity within the groups. We show that an efficient computational technique based on a generalized eige nvalue problem can be used to optimize this criterion. We have applied this approach to segmenting static images, as well as motion sequences, and fou nd the results to be very encouraging.