Precision Evaluation of RPCs in HRSI Orientation of Geoeye_1

Nowadays physical and abstract mathematical models are used with pushbroom satellite imagery, the imagery vendors do not provide ephemeris data (or ancillary data) for the end users, so one has to rely on mathematical sensor models for extracting 3D geospatial information. Rational functions are often applied in photogrammetric and remote sensing processing to represent the transformation between the image space and the object space whenever the rigorous model is made unavailable to the users intentionally or unintentionally. The RPC sensor model is considered to be a simple and effective way to approximate a rigorous camera model. However it has been seen that using RPC without the aid of ground control points contributes to geo positioning errors. This can be corresponding to terrain inherent interior and exterior orientation errors in the rigorous sensor model. These errors in geo positioning can be compensated by additional parameters in the image space that affect a translation of image coordinates. Therefore one needs to refine the supplied-RPC to achieve higher accuracy.This study reviews the developments in ground point determination from HRSI via the model of rational polynomial coefficients (RPCs). For geometrical accuracy improvement, the RPC sensor models by applying control points will be corrected.The potential for RPC block adjustment to yield sub-pixel geo positioning accuracy from HRSI is illustrated using results from experimental testing with two Geoeye_1 stereo images. Finally error propagation issues in RPC block adjustment of HRSI in various methods (supplied-RPC, supplied-RPC with compensated bias, generated RPCs with users) are considered.

Keywords:

HRSI Orientation , RPCs , Geoeye_1