Spatial size distributions: Applications to shape and texture analysis

This paper proposes new descriptors for binary and gray-scale images based on newly defined spatial size distributions (SSD). The main idea consists o f combining a granulometric analysis of the image with a comparison between the geometric covariograms for binary images or the auto-correlation funct ion for gray-scale images of the original image and its granulometric trans formation; the usual granulometric size distribution then arises as a parti cular case of this formulation. Examples are given to show that in those ca ses in which a finer description of the image is required, the more complex descriptors generated from the SSD could be advantageously used. It is als o shown that the new descriptors are probability distributions so their int uitive interpretation and properties can be appropriately studied from the probabilistic point of view. The usefulness of these descriptors in shape a nalysis is illustrated by some synthetic examples and their use in texture analysis is studied by doing an experiment of texture classification on a s tandard texture database. A comparison is perfomed among various cases of t he SSD and several former methods for texture classification in terms of pe rcentages of correct classification and the number of features used.