System and method for measuring image similarity based on semantic meaning

作者: Bernice Rogowitz , Aleksandra Mojsilovic , Jose Gomes

DOI:

关键词: Semantic similaritySemanticsMatching (graph theory)Set (abstract data type)Similarity (psychology)Artificial intelligencePattern recognitionPerceptionMathematicsImage (category theory)Meaning (philosophy of language)

摘要: A method includes deriving a plurality of semantic categories for representing important cues in images, where each category is modeled through combination perceptual features that define the semantics and discriminate from other categories; category, forming set comprising required frequently occurring features; comparing an image to said classifying as belonging one if all at least are present image. database contains information, information already classified network locations images documents containing images. The searched matching input query, comprising, e.g., image, text, or both.

参考文章(28)
Thomas Peter Minka, An image database browser that learns from user interaction Massachusetts Institute of Technology. ,(1996)
Frank Dellaert, Yanxi Liu, William E. Rothfusa, Classification Driven Semantic Based Medical Image Indexing and Retrieval ,(1999)
Jonathan I. Helfman, Mandala: An Architecture for Using Images to Access and Organize Web Informations Lecture Notes in Computer Science. pp. 163- 170 ,(1999) , 10.1007/3-540-48762-X_21
D. Comaniciu, P. Meer, D. Foran, Shape-based image indexing and retrieval for diagnostic pathology international conference on pattern recognition. ,vol. 1, pp. 902- 904 ,(1998) , 10.1109/ICPR.1998.711297
Bradley Horowitz, Jeffrey R. Bach, Chiao-fe Shu, Ramesh Jain, Amarnath Gupta, Charles E. Fuller, Threshold-based comparison ,(1997)