Abstract:
All remote sensing imagery, from satellites, is inherently subjected to geometric distortions.
Therefore geometric corrections, as preprocessing operations, are normally required prior to
imagery analysis and extraction of information.
This paper conducts geometric correction procedure of sample of raw satellite images using
georeferenced images (image-to-image registration) of the same area. In this procedure, many
well-distributed ground control points (GCPs) pairs (from both images) are identified. Then a
proper transformation polynomial is applied to map the original image GCPs coordinates into the
new georeferenced image GCPs coordinates. A resampling process is carried out to recalculate
the gray level values for pixels in the transformed output image (new pixels locations) based on
pixel values in the input image. Also, this paper presents an analysis study of the effect of
variation of the number of GCPs and the order of the mapping polynomials on the accuracy of
geometric correction process. The Root Mean Square Error (RMS), at the selected GCPs, are
calculated and used as a measure of accuracy of the obtained results.