A Multi Objective Genetic Algorithm based optimization of wavelet transform implementation

In this paper, an optimization method based on multi objective genetic algorithm (MOGA) is proposed in the design of a filter bank discrete wavelet transform (DWT). The different criteria are related to wavelet perfect reconstruction condition and minimization of the total number of non zero bits. Simulation results illustrate the optimization process applied to a case study : the Biorthogonal 9/7 DWT. They reveal improvement of compression performance and minimization of the number of non-zero bits. Performance measurements of the introduced method are even better than the original unquantized filter itself. Implementation results show a reducing hardware cost in comparison of proposed quantization method with trivial ones.