Adaptive Design fo Wavelets for OCntinuous and Discrete Wavelet Transforms to Determine Featrues in Signals

摘要

The wavelet transform in both its continuous and idscrete formulations has been investigatd as a transformation tool to represent information about phenomena in a signal as either feature in a surface or as dominant coefficients in a linear expansion ofthe signal. Althoutgh the utility of wavelet transforms for feature estimation is somewhat independent of specific details of the mother wavelet, the use of a data-derived wavelet, that is, one matched to the signal can improve the signal-to-noise ratio and resolution of features.