Content-based image retrieval via distributed databases

摘要: The overall objective of this paper is to present an extended application Content-Based Image Retrieval (CBIR) over distributed (decentralized) image databases. Traditional retrieval system design has implicitly relied on a local (centralized) query server, such as IBM's QBIC [1], Columbia's VisualSEEk [2], MIT's PhotoBook [3], and UCSD's Viagem™ [4]. With the growing popularity internet, however, focus research in area been shifted toward content Ng et al. [5] studied peer-clustering model for with assumption that collection at each peer node falls under one category. Even though, effective preliminary studies, it unable implant practical end-user behaviors. Lee [6] introduced novel approach study scenarios where multiple categories exist individual database storage network. This proven be method improve precision via identifying community neighborhood who shares similar collection. main behavior CBIR engine interactive environment. In proposed approach, sent all registered databases Response then collected transferred server supervised relevance identification applied identify final outcome search. quantified estimating statistical resemblance top candidates existing image. Comprehensive experiments demonstrate feasibility methodology.