A quad-tree approach to image segmentation which combines statistical and spatial information

摘要

A new approach to the problem of image segmentation is presented. By combining a nonparametric classifier, based on a clustering algorithm, with a quad-tree representation of the image, the scheme is both simple to implement and performs well, giving satisfactory results at signal-to-noise ratios well below 1. The results of an analysis of the algorithm are borne out by a comprehensive set of tests on Gaussian images and synthetic textures, which demonstrate its principal features.