基于内容的特大图像检索研究及实现  

A Project to Retrieve Super Large Image Based on Contents

作　　者：黄斌[1,2] 杨世洪[1,2] 吴钦章[1,2] 

机构地区：[1]中国科学院光电所,成都610209 [2]中国科学院研究生院,北京100049

出　　处：《计算机工程与应用》2005年第12期31-33,84,共4页Computer Engineering and Applications

基　　金：国家863信息获取主题"十五"期间支持项目:863-13大面阵彩色CCD数字航测像机系统研制(编号:2001AA131010)

摘　　要：如何实现特大图像基于内容的检索是航测、遥感和卫星图像等应用中的关键问题之一。主要从实现原理、算法流程和快速检索三方面讨论了这一问题,给出了以图像平滑缩小、自动分块、RGB亮度化和并行处理为主要技术的解决方案,最后介绍了一个基于上述设计原理的实例系统IPCBIR。It is one of the key problems for the application of aerial,remote sensing and satellite images how to retrieve super large image quickly.The paper discusses this from several aspects,including theory of realization,flows of arithmetic and fast retrieval and presents four primary techniques for this problem,which are downsizing smoothly,automatic partition,transforming RGB to brightness and concurrently processing.Finally,IPCBIR,an example developed by the authors according to the project is presented.

关 键 词：平滑缩小 自动分块 并行处理 基于内容的图像检索 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]