An efficient image retrieval system using Ordered Weighted Aggregation

Arslan S., YAZICI A.

1st International Symposium on Fuzzy and Rough Sets, Santa-Clara, Cuba, 5 - 08 December 2006, vol.224, pp.43-54 identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 224
  • City: Santa-Clara
  • Country: Cuba
  • Page Numbers: pp.43-54
  • Middle East Technical University Affiliated: Yes


In this study, an XML based content-based image retrieval system that combines three visual descriptors of MPEG-7(Dominant Color (DC), Color Layout (CL) and Edge Histogram (EH)) is introduced. The system is extended to support high dimensional indexing for efficient search and retrieval from a native XML-based DBMS. To do this, an index structure, called M-Tree, which uses Euclidean distance function for each feature, is used. In addition the Ordered Weighted Aggregation (OWA) operators are adapted for aggregating the distance functions of these features. The system supports nearest neighbor queries and various types of fuzzy queries; feature-based, image-based and color-based queries. The experimental results show that our system is effective in terms of retrieval efficiency.