3d Geometric Modelling of Ikonos Geo Images

Digital elevation model (DEM) extracted from IKONOS along-track stereo images with photogrammetric method is evaluated. As few as 12 GCPs are enough for the stereo photogrammetric bundle adjustment, which also filters the errors of the input data. With an area-based image matching users may produce high resolution DEMs with LE68 errors of 1 m to 4 m depending on the land covers. The best results (1.1 m-2.6 m) are obtained in bare soils, lakes, residential areas and sparse forests. The surface elevation of some of the areas (residential/ forests) did not affect too much the errors because the 1-2-storey houses in residential areas are sparse or because the images were acquired when there is no leave in the deciduous forests. An error evaluation as a function of the slope azimuths shows that the DEM error in sun-facing slopes is 1-m smaller than the DEM error in slopes away from the sun. 5-10 m contour lines could thus be derived with the highest topographic standard. 1. IKONOS STEREO DATA Three main attributes of IKONOS imagery for stereoscopic capabilities are 360o pointing capability, a base-toheight (B/H) ratio of 0.6 and greater, which is similar to aerial photographs, and the highest resolution available to civilian remote sensing and mapping communities. The 360o pointing capability enables the generation of across-track stereoscopy from two different orbits, such as with SPOT-HRV, as well as, along-track stereoscopy from the same orbit, such as with JERS-1’s Optical Sensor. The across-track solution has been used more since 1980; however the along-track solution as applied to space frame cameras has received renewed popularity in the past 10 years. In fact, same-date along-track stereo-data acquisition gives a strong advantage to multi-date across-track stereo-data acquisition because it reduces radiometric image variations (temporal changes, sun illumination, etc.), and thus increases the correlation success rate in any image matching process. This along-track solution to acquire stereo data is generally chosen by Space Imaging not only for scientific, but also for operational reasons. These stereo data are only available for governmental administrations as long as they are not used for commercial purposes (marketing, selling and distributing). Since Space Imaging does not provide the raw data with their ancillary data, preferred by the photogrammetrist community, only one quite similar to the GEO product can be ordered for stereo data. IKONOS stereo images are distributed in a quasi epipolar-geometry reference where only the elevation parallax in the scanner direction remains. For along-track stereoscopy with the IKONOS orbit, it approximately corresponds to a North-South direction, with few degrees in azimuth depending of the across-track component of the total collection angle. They are distributed in 8-bit or 11-bit GeoTiff format with an ASCII metadata file (including order parameters, source image and products file descriptions), however, detailed orbital information is not included. Since archive orders are generally not available for stereo-images, newly collected data is typically delivered in two or more weeks, depending upon order size, weather, and accuracy. ∗ Under contract with Consultants TGIS inc., 7667 Curé Clermont (Québec) H1K 1X2, Canada Largely extrapolated on results from similar systems mounted on aircraft platforms or from scanned aerial photos, IKONOS stereo-images should have a potential for creating DEMs with about 2m accuracy for use in national mapping (Ridley et al., 1997). This accuracy can be consistently achieved only if photogrammetric processing is employed (Li., 1998) and if the DEM is manually edited with 3D capability for surface elevation. Work still need to be carried out to evaluate the possibility for automating some processing steps and for using existing cartographic data, such as breaklines, hydrographic features and buildings. The objectives of this paper are to expand on these preliminary results with real IKONOS stereo images acquired from same orbit. Using a photogrammetric-based stereo-model developed at the Canada Centre for Remote Sensing (CCRS) (Toutin, 1995) and adapted to IKONOS images (Toutin and Cheng, 2000), the paper will evaluate DEM error when compared to ground truth, and track the error propagation from the input data to the final DEM. Different cartographic parameters affecting the accuracy are also evaluated. 2. PROCESSING OF IKONOS STEREO DATA The photogrammetric method of stereo data processing uses a parametric model that reflects the physical reality of the complete viewing geometry, and that corrects distortions due to the platform, sensor, Earth and deformations as a result of cartographic projection. Even though detailed sensor information for the IKONOS satellite is not released, such photogrammetric method was developed at CCRS for IKONOS stereo data using basic information available from the metadata and image files. For example, approximate sensor viewing angles can be computed using the nominal collection elevation and azimuth in addition with the nominal ground resolution. The CCRS model based upon principles related to orbitography, photogrammetry, geodesy and cartography was adapted for the specificity of IKONOS images. For stereo-images, both collinearity and coplanarity conditions are used to simultaneously compute the interior and exterior orientation parameters in a least-square bundle adjustment process (Toutin, 1995). The CCRS model has been previously applied with only a few ground control points (GCPs) (3 to 6) to VIR stereo data (SPOT, IRS, ASTER and KOMPSAT), as well as stereo SAR data (ERS and RADARSAT). Based on good quality GCPs, the DEM accuracy of this model was proven to be better than one pixel for medium-resolution VIR images, and one resolution cell for SAR images. The CCRS method is now fully ported into PCI OrthoEngine Satellite Edition V8.0 software. The software supports the reading of satellite data, metadata, GCP and tie points collection, bundle adjustment, orthorectification and mosaicking, stereo-model computation, image matching and DEM generation with either manual or automatic editing. After the stereo-model (colinearity and coplanarity equations) are computed using a minimum of six GCPs, an automated image matching procedure is used through a comparison of the respective grey values of the images (PCI, 2001). This procedure utilizes a hierarchical sub-pixel normalized crosscorrelation matching method to find the corresponding pixels in the left and right quasi-epipolar images. The difference in location between the images gives the disparity or parallax arising from the terrain relief, which is converted to X, Y, Z map co-ordinates using a 3D space intersection solution. Automatic and 3D-manual editing tools are finally used for the last step to improve DEM quality and coherency.