Research On The Storing And Organizing Structure Of Massive Remote Sensing Image Data

As information carrier with high capacity, remarkable reliability and great convenience, remote sensing image data plays an indispensable role in many fields like transportation management, land planning, military, resource, environment and disaster prevention. Via satellite, people can quickly get the high-resolution remote sensing images shot several days or even hours ago, increasing the timeliness and accuracy of obtained information.The data volume of remote sensing image is massive. Every day, there are large numbers of remote sensing images obtained by various means, and their number is increasing geometrically, up to Terabyte of Petabyte, in the wake of the improvement of image resolution. And the level of data volume is increasing fast and continuously with the passage of time, widening application of remote sensing and the development of satellite technology and sensor technology. Therefore, how to organize, store, manage and issue massive remote sensing image data effectively, in order for the remote sensing image data to serve the practical application faster and more accurately, has become an imperative problem.On the basis of above discussion, the distributed system of massive remote sensing data based on file system is proposed. Here are the major tasks of the thesis:1. The paper advances the standardized processing method and realization of remote sensing image data. In accordance with the characteristics of remote sensing image data and applicable demand for issuing and browsing remote sensing image data, its standardized processing approach is proposed, according to which the raw remote sensing mage data is processed. The process consists of: (1) data partitioning; (2) standardized nomination; (3) storage of standardized nominated data. The standardized remote sensing image data helps to speed the retrieval, transmission, and parallel computing; therefore, it promotes productive efficiency and contributes to the quick view and three-dimensional display of images.2. It proposes the retrieval algorithm and its realization. The algorithm, on the basis of customer's retrieval terms, is capable of constructing the target image storage route, according to which the computer searches, and returns with the retrieval result. The merit of this algorithm is the fast and accurate retrieval and little fluctuation of retrieval performance with the increase of data volume.3. It puts forward the SD-DRSDSS (Standard Data-Distributed Remote Sensing Data Storage System) and its realization. In line with standardized processing method of remote sensing image data, this paper proposes a distributed system which specializes in storing massive remote sensing image data. This system is simple in structure, easy to be managed and extended, and highly reliable; therefore, it is convenient for producing and issuing remote sensing products.4. It brings forward the retrieval algorithm and realization of SD-DRSDSS, in which fast retrieval can be realized in SD-DRSDSS system.