遥感分景数据的剖分索引模型  

Study on subdivision index model for scene-based RS data

作　　者：谭亚平[1,2] 程承旗[1] 耿晓晖[1] 

机构地区：[1]北京大学空天信息工程研究中心 [2]中国人民解放军61541部队

出　　处：《地理研究》2012年第6期1132-1142,共11页Geographical Research

基　　金：国家重点基础研究发展规划(973)项目(61399)

摘　　要：随着空间技术和对地观测技术的飞速发展,海量遥感影像索引成为目前国内外学术界和应用部门面临的一大难题,地球剖分系统的产生为这一难题的解决开辟了新的途径。本文首先研究了遥感分景数据的特征和EMD全球剖分的特征,然后针对现有遥感影像数据索引方法的不足,提出了基于EMD剖分系统的三种索引模型:基于影像—面片映射关系的索引模型、基于EMD数据模型的索引和剖分数据标识索引模型,从三个不同的方面探讨了解决索引问题的方法和思路。并给出了利用遥感分景数据的剖分索引模型对线状对象和面状对象进行索引应用的实例。遥感分景数据的剖分索引模型的建立为遥感影像数据的高效存储和索引提供了一种有益的新思路。With the rapid development of space technology and earth observation technology, modern remote sensing technology has been able to provide worldwide observed data collected in many means in a dynamic, fast and accurate manner. In this stage, remote sensing data increase rapidly accordingly, and how to manage and retrieve these data efficiently has become a major issue in both academic and application areas. Global Subdivision System opens up new avenues for solution to this problem. This peper studies the characteristics of scene-based remote sensing data and the features of the EMD Global Subdivision Model, and then in regard to the disadvantages of the existing index methods for remote sensing data, three index models are put forward based on the EMD Global Subdivision Model： an index model based on the mapping relationships of images and meshes, an index model based on EMD data model and an index model for global subdivision data identifiers. These three index models explore the solution to the issue in three aspects. And an application example is given to create the index of linear objects and planar objects by use of the scene-based RS data subdivision index model. This study mainly addresses the following aspects. （1） This research studies the characteristics of scene-based remote sensing data and the features of the EMD Global Subdivision Model, as well as the map- ping relationships and corresponding criteria between them. （2） In regard to the lack of space-time recording capabilities in the current storage and index methods for remote sens- ing data, an index model based on the mapping relationships of images and meshes is put forward. （3） In regard to the lack of support for multi-scale data of the current index models for remote sensing data, an index model based on EMD data model is established. （4） In regard to the lack of a unified index framework for multi-source remote sensing data, an index model for global subdivision data identifiers is put forward. （5） How to get remote s

关 键 词：索引模型 遥感影像 EMD模型 

分 类 号：P237[天文地球—摄影测量与遥感]