Region-based binary tree representation for image classification

Abstract

Image classification is a very challenging problem due to lack of effective representations. In this paper, a region-based binary tree representation incorporating with adaptive processing of data structures is proposed to address this problem. After an image is segmented, a binary tree is established to characterize its contents by using region merging method. Finally, an adaptive processing of data structure algorithm is employed to perform the classification task with binary tree representation. Experimental results on seven categories of scenery images show this region-based structural representation is superior to our previous work based on quadtree representation.