论文部分内容阅读
A framework of IIR two-channel bi-orthogonal filter banks is presented. The design of the analysis/synthesis systems reduces to the design of a single transfer function. The wavelet bases corresponding to the IIR maximally flat filters are generated. Use the wavelet bases in wavelet packets analysis, take information entropy function as information cost function, adjust the best bases dynamically, and the section of signal’s spectramfits its characteristics well. A lot of audio signals are processed. Compared to that of the wavelet transform and the wavelet packet transform using the Daubechies’ wavelet which has the same order with our IIR wavelet, the properties of our system are better
A framework of IIR two-channel bi-orthogonal filter banks is presented. The design of the analysis / synthesis systems reduces to the design of a single transfer function. The wavelet bases corresponding to the IIR maximally flat filters are generated. Use the wavelet bases in wavelet packets analysis, take the best bases dynamically, and the section of signal’s spectramfits its characteristics well. Compared to that of the wavelet transform and the wavelet packet transform using the Daubechies’ wavelet which has the same order with our IIR wavelet, the properties of our system are better