首页 | 本学科首页   官方微博 | 高级检索  
     

一种基于聚类的对等网络图像搜索机制
引用本文:张全信,胡峰松,欧阳竟成. 一种基于聚类的对等网络图像搜索机制[J]. 计算机应用, 2006, 26(3): 675-0678
作者姓名:张全信  胡峰松  欧阳竟成
作者单位:湖南大学,计算机与通信学院,湖南,长沙,410082;湖南大学,计算机与通信学院,湖南,长沙,410082;湖南大学,计算机与通信学院,湖南,长沙,410082
基金项目:教育部高校骨干教师资助计划
摘    要:将基于内容的图像检索应用到P2P网络中,提出了一种基于“类簇”的P2P网络信息搜索机制。结合聚类方法和Small world原理,将相似的图像文件聚集到同一个“类簇”中,不同簇之间建立远距离连接;查询时先定位到所属的簇,在簇内广播查询,并根据查询记录更新节点路由表。仿真实验表明该搜索机制具有良好的搜索性能。

关 键 词:对等网络  图像检索  聚类  路由表更新
文章编号:1001-9081(2006)03-0675-04
收稿时间:2005-09-26
修稿时间:2005-09-26

Clustering-based image retrieval mechanism on P2P network
ZHANG Quan-xin,HU Feng-song,OUYANG Jing-cheng. Clustering-based image retrieval mechanism on P2P network[J]. Journal of Computer Applications, 2006, 26(3): 675-0678
Authors:ZHANG Quan-xin  HU Feng-song  OUYANG Jing-cheng
Affiliation:School of Computer and Communications, Hunan University, Changsha Hunan 410082, China
Abstract:The content-based image retrieval system was introduced into P2P network and a novel search model based on "cluster clustering" was presented. Firstly, according to clustering algorithm and the small-world theory, image files with similar feature were collected together while dissimilar clusters were connected by shortcuts. Furthermore, a query reached its target cluster, and was broadcasted in the cluster. Founded on the previous query result, routing tables were updated. Simulation results show that the model is efficient.
Keywords:peer-to-peer network   image retrieval   clustering   updating routing table
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《计算机应用》浏览原始摘要信息
点击此处可从《计算机应用》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号