Linear and nonlinear interpolation for 2-dimensional image enhancement

Abstract

The problem of enhancement of two dimensional images by recursive interpolation is considered. A nonlinear interpolator is modelled by considering the constraint that the observations of the image by a scanning device are always positive. Both linear and non-linear interpolators are implemented on a two dimensional picture when the observations are corrupted by noise. Dimensionality problems associated with interpolation are discussed and experimental results are presented.