A Framework for semantic modeling of images

Abstract

In this paper we propose a framework for semantic modeling of images. The framework includes components for extraction of low-level image features, for content-based image retrieval and methods for incorporation of semantic knowledge into the retrieval process.
The semantic information is represented by semantic association rules, which are used for an interactive annotation of the image data. So, the paper approaches modalities for reducing the semantic gap between the low-level characteristics automatically extracted from the visual content and the high-level concepts. The experiments through the framework were realized on collections of images from nature and medical domain.