Image domain ontology fusion approach using multi-level inference mechanism

One of the main challenges in content-based or semantic image retrieval is still to bridge the gap between low-level features and semantic information. In this paper, An approach is presented using integrated multi-level image features in ontology fusion construction by a fusion framework, which based on the latent semantic analysis. The proposed method promotes images ontology fusion efficiently and broadens the application fields of image ontology retrieval system. The relevant experiment shows that this method ameliorates the problem, such as too many redundant data and relations, in the traditional ontology system construction, as well as improves the performance of semantic images retrieval.