Rational function model-based image matching for digital elevation models

This paper presents an image matching technique for IKONOS satellite imagery based on rational function models (RFMs). This algorithm adopts the object-space approach and reduces the search space to within the confined line-shaped area called the piecewise matching line (PML). Also, the detailed procedure of generating 3D surface information using the RFM coefficients (RFCs) is introduced to illustrate an end-user's point of view. The final digital elevation model (DEM) generated using the proposed scheme shows a mean error of 2·2 m and rmse of 3·8 m compared with that from a 1:5000 scale digital map.