Self-Organization in Wavelet-Based Hidden Markov Model Signal Processing
Miloš Milovanović, Milan Rajković · arXiv (Cornell University) · 2012
A model of self-organization in the wavelet signal processing is developed to provide criterion for the choice of the optimal wavelet basis. At the same time the model quantifies self-organization and, with the optimal basis, performs superior denoising with respect to the standard denoising methods based on the wavelet hidden Markov model. The method is presented for the one-dimensional signals and may be easily generalized to higher dimensions.