Measurement of signal similarity using the maxima of the wavelet transform

The local maxima of the absolute values of the wavelet coefficients are used to measure the cross-correlation between a pair of input signals. Unlike the classical method which uses the cross-correlation coefficients where all data points are given equal weights, this approach utilizes the edge points only. These edge points are compared in pairs forming a cross-correlation vector whose length is equal to the number of scale levels in the wavelet transform. An algorithm which systematically matches the edge points provided by the maxima of the wavelet coefficient modules is presented. Methods for selecting the wavelet and neighborhood parameter are also discussed.<>