Signal analysis via wave-splitting based on the wavelet transform

The discrete dyadic wavelet transform is modified so that, at each scale, the waveform is split into several channels. These channels correspond to decomposition of information in such a way that signal analysis and processing may be easily performed. It is shown that the split dyadic wavelet transform forms a complete representation. Application of this technique is demonstrated through an example.<>