Abstract for presentation at The 13th Australasian Remote Sensing and Photogrammetry Conference

Successful exploitation of the high accuracy potential of High Resolution images depends on mathematical modeling of their imaging sensor. Unfortunately some companies do not intend to release ancillary data and Raw images. Therefore Rigorous Models can not be used easily to extract precise 3D spatial information from these images. Recently, several 2D and 3D approaches have been reported by various investigators to tackle this issue. Amount these approaches, Rational Function Model (RFM) is the most considerable one which has been investigated by photogrammetrists committee. RFM has some advantages such as being independence of orbital parameter and image geometry. However, it’s not an exact method and has some imperfections such as high sensitivity to input errors, and GCPs. Rigorous methods such as Orbital Parameter Model based on collinearity equations, model precisely the geometry of image at the time of imaging. And consequently needs less GCPs comparing to RFM. However the problem of loosing the imaging geometry at the time of imaging has to be considered. This paper offers a model to solve this problem. The developed model has been tested successfully on SPOT L1B, and IKONOS Geo images (orbital parameters of IKONOS satellite have been estimated). This is followed by the results of various geometric accuracy tests carried out using different parameters and combination of control and check points. For the IKONOS Geo the accuracy of 0.79 m and for the SPOT L1B, 10.96 m accuracy was achieved using the Orbital Parameter Model. Later the results of the developed model have been compared with RFM results (in two state of with and without denominators).