DERIVING 3-D SHORELINES FROM HIGH RESOLUTION IKONOS SATELLITE IMAGES WITH RATIONAL FUNCTIONS

Rational Functions (RFs) have been used in the remote sensing community to replace the rigorous sensor models that are sometimes confidential or used to achieve a greater processing speed. In particular, they are used to define the imaging geometry for the data from emerging high-resolution satellites, allowing vendors to keep the sensor models confidential. This paper presents methods for and results from deriving 3-D shorelines in the Lake Erie area from simulated high-resolution IKONOS stereo images based on rational functions. The solutions of photogrammetric intersections based on upward and downward RFs are examined. The upward RFs project from the ground space to the image space, while the downward RFs project from the image space to the ground space. The two solutions are both based on using Taylor’s theorem to linearize the nonlinear rational function and computing the ground positions iteratively. The shorelines derived from the simulated IKONOS images are compared with those obtained from existing maps and NOAA stereo aerial photographs. The issues concerning production of tidecoordinated shorelines are also discussed.