Survey of resampling techniques using MSS and synthetic imagery

The objective of this survey is to investigate the methods of interpolation and deconvolution for image restoration The methods evaluated are nearest neighbor, bilinear interpolation, cubic convolution, and two-dimensional deconvolution. The effects of these restoration methods are demonstrated using Landsat multispectral scanner (MSS) data and synthetic imagery.

The effect of these restoration methods are compared as to resolution and spatial frequency effects. The edge effect, a situation that occurs when fill (non-image) data is interpolated with image data, is also addressed.