New image retrieval principle: Image mining & visual ontology

Abstract : Image data are omnipresent for various applications. A considerable volume of data is produced and we need to develop tools to efficiently retrieve relevant information. Image mining is a new and challenging research field which tries to overcome some limitations reached by content-based image retrieval. Image mining deals with making associations between images from large database and presenting a resumed view. After a state of the art in the image retrieval field, this chapter presents some work and ideas about the need to define new descriptors to integrate image semantics. Clustering and characterization rules are combined to reduce the research space and produce a resumed view of an annotated image database. These data mining techniques are performed separately on visual descriptors and textual information (annotations, keywords, web pages). A visual ontology is derived from the textual part, and enriched with representative images associated to each concept of the ontology. Ontology-based navigation can also be seen as a user-friendly and powerful tool to retrieve relevant information. These two approaches should make the exploitation and the exploration of a large image database easier.
Document type :
Book sections
Complete list of metadatas

https://hal-supelec.archives-ouvertes.fr/hal-00260998
Contributor : Evelyne Faivre <>
Submitted on : Thursday, March 6, 2008 - 9:16:54 AM
Last modification on : Thursday, March 29, 2018 - 11:06:03 AM

Identifiers

  • HAL Id : hal-00260998, version 1

Collections

Citation

Marinette Bouet, Marie-Aude Aufaure. New image retrieval principle: Image mining & visual ontology. Multimedia Data Mining and Knowledge Discovery, Springer, pp.177-196, 2006. ⟨hal-00260998⟩

Share

Metrics

Record views

137