The Research On Geometry Processing System For Remote Sensing Image Based On Five Layers Fifteen Levels Data Organization Structure

With domestic satellite launching successfully the remote sensing image products in many industry play a more and more important role in many industry. But the image to be affected by many factors to make the original satellite images with more or less distortion, which needs to eliminate the distortion through a certain method in order to improve the quality of the images, so that the subsequent production and analysis of remote sensing product. However, the current image geometric processing mainly depends on the commercial software. The process based on the geometry of commercial software need input image data and parameters, can’t form a complete set of geometric process and meet the needs of business processing. So the construction of geometry processing system of domestic high satellite imaging model, which achieves domestic satellite images automatically and accurately and fast geometric processing system, has an extremely important practical significance.Base on the five layers fifteen levels data organization structure proposed Institute of Remote sensing and Digital Earth Chinese Academy of Sciences and the general steps of the traditional geometric processing, the paper studies the process of remote sensing image geometric processing quickly and in business. The research contents of this paper are as follows:1) By analyzing each process characteristics in remote sensing image geometric processing and combing with the advantages of new data organization structure of five layers fifteen levels, this paper adds the segment method based on the five layers of fifteen levels segmentation standard to the geometry processing process. After processing, the images will be stored in the form of 1000 * 1000 tiles. The geometric processing as early step of remote sensing products production provides image parallel processing and distribution of the late storage with the basis.2) This paper researches the method of structuring the clear sky background field of the cloud detection module in the geometry processing. Based on the tiles from different time in comprehensive database the clear background file will be structured. As the difference of detection of time and area, the clear background field will be different.Threshold value will be dynamic based on this background filed.The accuracy of cloud detection will be improved base on this threshold value3) On the basis of the five layers fifteen segmentation standard, this paper proposes a new method of structure control point image database. The images used in image registration are segmented in accordance with the five layers of fifteen levels segmentation standard will be stored in file system in a form of tile information combined with its path information, which can realize the rapid retrieval automatic image matching. The paper also provides automatic control point image storage method.4) Based on the demand of Industry for geometry processing system, the paper designs the system architecture, order processing mode and the interface to apply to the demand of the industrial integration of remote sensing image processing,so that remote sensing products better service in various industries.