期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2008
卷号：XXXVII Part B1
页码：913-918
出版社：Copernicus Publications
摘要：The DEM (Digital Elevation Model) and ortho-image have been used in various fields such as 3D (dimension) map, 3D urban model, simulation, and so on. Today, users require more detail data but many existing DEMs and ortho-images have low spatial resolution. Meanwhile, high resolution satellite images such as QuickBird, IKONOS, KOMPSAT-2 images have appeared recently, so we propose the technique for automated update of existing DEMs and ortho-images database using high resolution satellite images. This method uses stereo images matching technique by GCPs (Ground Control Points) for DEM generation from satellite images. We automatically extracted GCPs from high resolution satellite images by matching satellite image against existing ortho-image and using DTED (Digital Terrain Elevation Data) Level 2 data (30m spatial resolution). We used the robust estimation, RANSAC (Random Sample Consensus), to minimize error of extracted GCPs. The tie-points for automatic stereo matching is extracted from stereo pairs using existing DEMs. We tried to generate accurate DEMs, which is DTED Level 3 (10m spatial resolution). We also made ortho-images from high resolution satellite images by existing DEMs and GCPs generated automatically. We applied our method to IKONOS, SPOT 5, QuickBird for test. The test result showed that automatic generation of DEMs and ortho-images was performed successfully. We could update DEM from DTED Level 2 to DTED Level 3 and we could update existing ortho-images using high resolution satellite images
