节点文献

基于主色和主色块的图象检索

Image Retrieval Based on Dominant Color and Dominant Color Block

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 袁昕朱淼良

【Author】 Yuan Xin, Zhu Miaoliang (Artificial Intelligence Institute of Zhejiang University, Hangzhou 310027)

【机构】 浙江大学人工智能研究所!杭州310027

【摘要】 基于内容的图象检索CBIR(Content Based Image Retrieval)是多媒体信息管理系统的重要组成部分.本文在简要介绍CBIR技术的基础上,提出了用于CBIR系统新的图象内容的特征提取的方法和基于该特征的图帘匹配方法--基于主色的图象特征提取(DCE,Dominant Color Extracting)和匹配方法(DCM,Dominant Color Matching),并在此基础上设计并实现了WWW发布方式的SPRING实验系统,取得了理想的实验结果。

【Abstract】 Content Based Image Retrieval (CBIR), which is an important component of Content Based Retrieval (CBR) system, is a new image retrieval technology based on image content. Based on the introduction of the technology of CBIR systems, the function modules of system and the new method of feature extracting and matching of image is summarized in this paper. This method is called Dominant Color Block (DCB) method. A 3-D clustering technology and statistics calculation based Dominant Color Extraction (DCE) method and Dominant Color Matching (DCM) method is also realized in this paper. On the basis of the detail study or CBIR technologies, a CBIR system named SPRING supporting WWW distribution is designed and implemented, which shows a good performance.

【基金】 国家自然科学基金(No69875015);浙江省自然科学基金(111101 N29809).
  • 【文献出处】 计算机工程 ,Computer Engineering , 编辑部邮箱 ,2000年S1期
  • 【分类号】TP393.09
  • 【被引频次】9
  • 【下载频次】135
节点文献中: 

本文链接的文献网络图示:

本文的引文网络