作者: James Wang , Jia Li
DOI:
关键词: Measure (data warehouse) 、 Image (mathematics) 、 Hidden Markov model 、 Image based 、 Stochastic process 、 Linguistics 、 Pattern recognition 、 Artificial intelligence 、 Search engine indexing 、 Statistical model 、 Computer science 、 Association (object-oriented programming)
摘要: The present invention provides a statistical modeling approach to automatic linguistic indexing of photographic images. uses categorized images train dictionary hundreds models each representing concept. Images any given concept are regarded as instances stochastic process that characterizes the To measure extent association between an image and textual description associated with predefined concept, likelihood occurrence based on characterizing is computed. A high indicates strong image. utilizes two-dimensional multi-resolution hidden Markov demonstrate accuracy potential in