Complementary filter banks for image compression

Abstract

In this paper we present a nonseparable complementary filter (CF) bank for image compression. We base our design of the filter bank on the two-dimensional (2D) multirate processing and on properties of complementary filters. Perfect reconstruction is guaranteed when the CF bank is implemented with quincunx sampling and filtering, but not otherwise. In all cases, however the analysis and synthesis are performed without aliasing. The CF bank can be related to wavelet theory, permitting us to derive scale and wavelet functions for different sampling and filters. Examples illustrating the performance of the CF banks in image data compression are given.