The Analysis And Design Of Remote Sensing Imagery Information System

In the remote sensing application area such as environment monitoring, change detection, and military reconnaissance etc, remote sensing imagery information system is the most important fundamental infrastructure to manage and exploit multi-temporal, multi-spectrum, multi-resolution and multi-source remote sensing image. This paper put the emphasis on several key techniques of such system and the main work and originalities are as follows:1 .Techiniques for service information on Web, the characteristic of Web-based application and traditional application and also the general Web database architecture are thoroughly analyzed to make Web-based software architecture re-use possible.2.The object-oriented modeling process for architecture-driven software engineering is introduced into the software analysis and designation, a general, distributed remote sensing image information system architecture model is presented with typical three hierarchical computing architecture.3.A hierarchical object data model of remote sensing image is presented, and in order to use commercial relational database system to implement data management, the method for wrapping objects to relational schema is discussed.4.Based on the analysis of the research of image metadata, the metadata system of remote sensing image information system is designed and proposed.5.The technique of very large volume data storage system designation is elaborated. Then, the abstract model and also the evaluation method of storage system are proposed and a specific distributed storage system is introduced.6.The decorrelation performance, boundary problem, scale coefficients process and statistic characteristics of different integer wavelet transform are studied experimentally.7.Two type of lossless image compression scheme are investigated and improved in different extends. Then, a new efficient wavelet-based lossless compress scheme, which can more satisfy the requirements of network-based application of large RS image, is developed. The result is comparable to the state of art predicate technique based scheme JPEG-LS.8.A set algebraic model that can uni1~ the keywords-based relation query and content-based retrieval is presented. The GIS and perceptual feature supported remote sensing image CBIR techniques are studied theoretically and experimentally.9.The algebraic operator of image query language are summarized and classified, and then a primary image query language is established.I O.Finally, a practical Web-based remote sensing image information systemæ¡°mage Map Library is developed based on the research work of the thesis.