Using multidimensional scaling technique in image dimension reduction for satellite image

Abstract

A satellite multispectral sensor provides data in the form of several spectral image of particular area of the earth under observation. The amount of data by multispectral exhibit high-inter band correlation with redundancy of the information. In this research, we suggested using multidimensional scaling technique as a statistical technique to reduce image dimensions for reconstruction of a new image form multispectral image. The results proved the efficiency of this technique in providing high quality low dimension images based on the value of Peak Signal to Noise Ratio (PSNR) that was measured to the new image. At the same time, the ability of this technique to reduce dimensions while preserving the basic characteristics of the image has been confirmed.