Image Clustering using Semantic Tolerance Relation Model

Y. Dai and D. Cai (Japan)


Semantic categorization, semantic tolerance relation model, image clustering algorithm


Because the nature of the concepts regarding images in many domains are imprecise and can be overlapped, and also there is the gap between the ability of machine computation and the high level concepts of human perception, in this paper, we define the tolerance degree between semantic categories, and present a method of semantic-based image presentation using the tolerance relation model. We apply the proposed method to construct a tolerance relation-based semantic clustering algorithm of images. The decision rules, for the semantic categorization based on the concepts of nature vs. man made, human vs. non-human, and the both of them, are given. The performance evaluations of the proposed methods are executed.

Important Links:

Go Back