The relationship between matched filtering and wavelet analysis is discussed. A wavelet synthesis procedure is offered, which relies on the a priori information on the signal shape (apparatus funñtion) and noise correlation function. This allows adaptation of the wavelet analysis of measurement data for maximal signal denoising. The results of modeling based on the Gaussian signal model are given.