Comparative Analysis of Various Methods for Preprocessing of Satellite Imagery

The images obtained by remote sensing systems are not quite often sufficient for high precision applications due to various distortions. The distortions can be divided into three layersGeometric, Radiometric and Atmospheric distortions/Errors. The first layer deals with the steps to be carried out for correcting the Geometric distortions/errors through generic non-parametric models that relates remotely sensed image to terrain coordinates. The second layer is dealt about the five methods used for Relative Radiometric Correction (RRC) such as Pseudo Invariant Features (PIF), RadiometricControl Set (RCS), Image Regression (IR), No-Change set determined from scatter grams (NC) and Histogram Matching (HM) and the third layer deals with an Atmospheric Correction which is to retrievethe surface reflectance from remotely sensed imagery by removing the atmospheric effects by using theoretical model called look-up table. A comparative study of the various mathematical techniques hasbeen discussed in this paper for the betterment of preprocessing of satellite images.