Image magnification using families of fast orthogonal transforms

We outline a method of image magnification based on parametric families of fast orthogonal transforms which arise from the multiresolutionlwavelet paradigm of Mallat and Meyer. The essential idea is to consider the original image as a low-frequency component of a family of potential enlargements, all of which are computable in real time. We sketch the underlying theory and pertinent algorithms. Several specific examples are detailed.