An image retrieval model based on fuzzy triples

This paper proposes an image retrieval model for in dexing and retrieving images with fuzzy triples. Th e well known triple framework based on 2 dimensional (2D) strings is a novel index mechanism in that it allows us to simply specify the spatial structure o f images, guaranteeing fast retrieval time. However , there are two significant drawbacks in this framewo rk; one is that it can not accommodate a concept-based image retrieval and the other is that it does not deal with an inexact match among directions. T he former is crucial when answers turn out to be relev ant only when they are conceptually related with us er queries and the latter is useful especially when th e spatial relationship can not be specified in term s of eight directions. Our model supports the concept-ba sed image retrieval as well as the inexact match wi th a fuzzy triple matching performed when evaluating que ries. To demonstrate the feasibility of the model, we also develop a prototypal system directly supporting it.