An extendible semantic-based content representation method for distributed image databases

This paper introduces a scheme for semantic-based image content representation in a distributed heterogeneous database environment. The summary-schemas model (SSM) is used as the underlying platform. With the ability of summarizing the content information and guiding data distribution, the SSM provides a quality-guaranteed and time-efficient image accessing strategy. The advantages of representing and indexing images on the platform of SSM are as follows: (1) the SSM has mathematical capability of representing semantic content precisely and concisely; (2) content-based queries are efficiently resolved through hierarchical organization and classification of image objects according to their contents; and (3) imprecise text descriptions can be included in queries, which further enhances the searching capability of image databases. A conceptual framework of the proposed model is presented. The proposed scheme has been simulated, and the results are presented and analyzed. Finally, the proposed image representation and retrieval scheme has been compared and contrasted against several content-based retrieval systems as presented in the literature.