基于内容检索的图像系统的关键技术及其实现

The Content－based Image Retrieval System＇s Key Technologies and its implementation

大量多媒体化信息的产生使得基于内容的检索技术成为研究的热点.该文先给出了基于内容的图像检索系统的系统结构.然后归纳并详细阐述了该系统所包含的5种主要关键技术,提出要实现趋向语义的图像查询系统应建立在相关性反馈、多种特征组合查询和对图像内容加注的文字进行全文本查询相结合的基础上,并且在原型系统上给出了实验结果.

： The creation of large multimedia information makes the content－based retrieval technology become a research focus.We first discuss the architecture of content－based system,from which we induce and discuss in details the five key technologies of such system.And we point out that establishing the semantic－based image retrieval should be based on the combination of the relevance feedback,multiple characters combined retrieval and metadata full－text retrieval technology.Finally the experimental results on our prototyping system are also given．