EFFECT ON DIGITAL IMAGE CLASSIFICATION DUE TO PIXEL RESAMPLING

Abstract

As a wide variety of satellite images with different resolutions become available due to the launch of many remote sensing satellites, techniques of extracting necessary information by integrating these data are increasingly important. And as the development of GIS data and remote sensing image data are becoming very common. GIS is now an indispensable system as a platform of integrating information. So while integrating the image data from one system to another system, the process of resampling comes to picture. The three commonly used resampling techniques in practice are Nearest Neighbor, Bilinear Interpolation and Cubic Convolution. As the pixel values are altered due to resampling there may be a significant effect on digital image classification process. In this present study an attempt has been made to highlight the effect on digital image classification due to pixel resampling. For this purpose, the work has been planned in two stages. In the first stage, difference images have been generated between the supervised classified raw and various resampled images, to know the effect of resampling on image classification. In the second stage, a comparison of various supervised classified resampled images and subpixel classification (proposed by Applied Analysis Incorporation) of the raw image is done.