Adaptive multiwavelet initialization

摘要: The multiwavelet concept uses a set of scaling functions and wavelets to generate an orthogonal multichannel multiresolution pyramid decomposition for finite energy signals. multiple extract the lowpass information input data set, multiwavelets bandpass information. When filters associated with are cascaded, decomposition/reconstruction system is formed each convolution operator having several inputs outputs. However, there only one available initialize this process. This paper addresses question how best modify using prefilters so that its contains most useful chosen application. "best" determined by minimization preselected components. resulting is, therefore, signal adaptive, under appropriate conditions, perfect reconstruction can be achieved proper postfiltering. detailed discussions in focussed on two-wavelet two-scaling function case; general case addressed at end paper. A compression example provided demonstrate performance optimally initialized method compare it single wavelet another initialization proposed other authors.