Optimization Design of Biorthogonal Filter Banks for Image Compressing

A new kind of 9/7 wavelet basis, the optimum biorthogonal wavelet basis, is establishedd according to the effect of image compression by adopting the SPIHT algorithm based on the idea of biorthogonal wavelet filter bank with the lattice structure in this paper. Lenna and the Goldhill were adopted as the standard training images; and the maximal peak signal-to-noise values were obtained with the compression rate as the parameter. So, the looking for the optimal biorthogonal wavelet basis is simple to the optimization problem. The genetic algorithm was adopted here. It is illustrated compression effect of other standard image by finding optimum biorthogonal wavelet basis, that is, comparing compression result with Antonini´s 9/7 wavelet basis. This method is feasible in the experiments, and the inspection is effective.