چکیده:
In the high resolution satellite images (HRSI), the high accuracy depends on accurate
mathematical models for the satellite sensors. Because, there is not satellite orbit information
for the most of the new HRSI, this issue is very important in geometric correction of satellite
imageries. The pre-processing of satellite images consists of geometric and radiometric
characteristics analysis. By performing these operations, it is possible to correct image
distortion and improve the image quality and readability. The radiometric analysis refers to
mainly the atmosphere effect and its corresponding to feature's reflection, while the geometric
correction refers to the image geometry with respect to sensor system with the launch of
various commercial high-resolution earth observation systems. However, during the
generating of satellite imageries, the projection, the tilt angle, the scanner, the atmosphere
condition, the earth curvature and etc., will cause the satellite images to have distortion. So,
It is necessary to correction these distortions before one can really use it as a precise
measurement in the large scale operations. In this paper, different non-rigorous (generic)
mathematical models investigate for geometric corrections over an IRS P6 (Resourcesat-I)
Satellite imageries (exactly LISS IV sensor images) in Iran. The LISS IV sensor of the IRS-P6
(Resourcesat-I) satellite has the spatial resolution 5.8 m with a enhanced spectral resolution.
These different geometric models for performing the geometric correction on the satellite
imageries includes Rational function models, different orders of polynomials models,
projective, affine ( 2D and 3D) and DLT ( Direct Linear Transformation) model with the
different numbers of GCP points. Therefore, these mathematical geometric models can be
applied to determine the ground point coordinates in object space and so can be used to
provide good sufficient insight about the rectified images. In fact non-rigorous mathematical
models for geometric corrections of any images can be defined as the models, which can be
precisely, present the relationship between the image space and the object space. With
implementation of different transformation models on the test data in IRAN, we found the best
transformation model in geometric corrections which have the minimum RMSE (Root Mean
Square Error) rather than another transformation models