State-of-the-art image retrieval techniques have been developed to support high-level (semantics) querying and browsing of images. For certain user queries, it is also very important to know about the information about the remote image databases on the Internet so that the queries can be guided toward the most relevant databases. In this paper, we introduce a novel system, Sem View, which supports both visual and semantic queries and intelligently ranks the distributed image databases for the queries. Our system summarizes the visual and semantic contents of each database in a metadatabase, which is a collection of automatically generated visual, semantic and statistical metadata. With the metadatabase, our system can effectively guide the user queries toward the most relevant image databases in a distributed environment for better retrieval performance.
[1]
Lei Zhu,et al.
Keyblock: an approach for content-based image retrieval
,
2000,
ACM Multimedia.
[2]
Yimin Wu,et al.
A feature re-weighting approach for relevance feedback in image retrieval
,
2002,
Proceedings. International Conference on Image Processing.
[3]
Aidong Zhang,et al.
Automatic Annotation and Retrieval of Images
,
2004,
World Wide Web.
[4]
David M. Mark,et al.
Webview: A Distributed Geographical Image Retrieval System
,
2002,
DG.O.
[5]
Aidong Zhang,et al.
Data Resource Selection in Distributed Visual Information Systems
,
1998,
IEEE Trans. Knowl. Data Eng..
[6]
Aidong Zhang,et al.
Semantics Retrieval by Content and Context of Image Regions
,
2002
.