Technique for designing biorthogonal wavelet filters with an application to image compression

A technique for designing new symmetric biorthogonal wavelets from a given symmetric regular filter is presented. The main idea is to find a symmetric complementary filter of a given regular filter such that it has the least mean square (LMS) amplitude deviation from the ideal halfband lowpass filter. New biorthogonal wavelet filter pairs can be obtained via factoring the product of the complementary polynomial and the given binomial. By applying these new symmetric biorthogonal wavelet filters to the compression of some complicated images an improved result in reducing artefacts may be achieved