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：2004
卷号：XXXV Part B3
页码：1024-1029
出版社：Copernicus Publications
摘要：Image resampling according to epipolar geometry is a prerequisite for a variety of photogrammetric tasks such as image matching, DEM and ortho-photo generation, aerial triangulation, map compilation, and stereoscopic viewing. The resampling process of imagery captured by frame camera has been established and implemented in current Digital Photogrammetric Workstations (DPW). Scanning analogue images or directly using digital cameras can produce digital images, which are the input media for DPW. So far, there is no digital frame camera capable of producing radiometric and geometric resolutions that are comparable to those associated with analogue ones. To overcome such a limitation, linear array scanners have been developed to capture scenes through multiple exposures of few scan lines along the focal plane. This imaging scenario makes the perspective geometry of line cameras more complicated than that of frame images. Moreover, established procedures for resampling frame images according to epipolar geometry are not suitable for scenes captured by linear array scanners. In this paper, the geometry of epipolar lines in scenes captured by linear array scanners moving with constant velocity and constant attitude is analyzed. The choice of this model is motivated by the fact that many scanners can be assumed to follow such trajectory during the short duration of scene capture (especially when considering space borne imaging platforms). The paper starts by establishing the mathematical model describing the shape of epipolar lines in the captured scenes from a platform moving with constant velocity and constant attitude. The paper proceeds by presenting simulated experimental results using scenes captured according to SPOT, IKONOS and three-line camera imaging configurations, where we show the effect of different stereo-coverage methodologies (i.e., tilting the sensor across the track direction – SPOT, tilting the sensor along track – IKONOS, and using three linear array scanners) on the shape of the epipolar lines. For these scenarios, a measure is introduced to quantify the deviation of the resulting epipolar lines from straightness. Small deviations from straightness will allow for better capability of resampling the captured scenes according to epipolar geometry and thus lead to more straightforward subsequent processing (e.g., DEM generation, stereo-viewing, etc.).
