POLYNOMIAL PRESERVING ALGORITHM FOR DIGITAL IMAGE INTERPOLATION

Interpolation is a process of estimating the intermediate values of th e samples from their neighbouring points. The original samples and the new interpolation values can be regarded as realization of a given si gnal at different resolution levels. An interpolation algorithm is the refore served as a link between different resolution levels of the sig nal. In this paper, a family of iterative interpolation algorithm is i ntroduced. The algorithm uses splines iteratively and preserves certai n polynomials. Comparison with cubic convolution, cubic spline, Daubec hies' wavelet and FFT-based interpolations is made. The tensor product of two one-dimensional interpolations is applied to digital images. ( C) 1998 Elsevier Science B.V. All rights reserved.