摘要
采用移动多Agent技术,提出了一种灵活的基于内容的W eb图像检索系统模型。系统将传统的W eb图像检索功能封装于不同的Agent中。各Agent自主运行并相互协调,共同完成分布式W eb图像检索。由于图像搜索策略封装于移动Agent中,可在数据源本地实现快速的图像匹配,减少了网络信息交换;网络管理Agent消除了异构网络带来的通信障碍。实验表明该图像检索系统具有良好的自治性、伸缩性和适应性以及灵活的在线扩展能力。
Based on the concept of mobile multi-agent, a novel flexible Content-Based Image Retrieval （CBIR） system is proposed for Web image retrieval. Traditional CBIR functions are encapsulated in various agents. These agents run autonomously and cooperate harmoniously with each other. Their collective coordination results in efficient retrieval of Web images. The mobility of image search agents can minimize the network traffic by performing fast local similarity match at the database source. Furthermore, network management agent can eliminate the heterogeneity of networks. The implementation validates its appealing properties such as preferable autonomy, scalability, adaptivity as well as on-line extension.
出处
《计算机应用研究》
CSCD
北大核心
2005年第11期189-192,共4页
Application Research of Computers
基金
铁道部科技研究开发基金资助项目(2003X040-A)
西南交通大学科学基金资助项目(2002A04)
