A resolution-independent image representation for digital cinema

Abstract

We present a generic architecture for a novel type of image and movie representation designed to allow resolution-independent manipulation of movie sequences. By expressing the input signal as a sum of analytically-defined basis functions across space, time and scale, it becomes possible to compute image values at any location in the sequence, thereby providing a basis for both spatial and temporal resolution-independence. Distortion-constrained encoding of the basis function coefficients is furthermore built into the architecture to achieve data compression whilst minimizing signal degradation.