Modified wavelets and their applications
L. V. Novikov · Journal of Communications Technology and Electronics · 2006
Wavelets obtained from known orthonormal wavelets modified by the impulse response of a stationary linear system are proposed. It is shown that the new wavelets offer additional possibilities for signal processing in the presence of noise. In particular, these wavelets provide for estimation of linearly transformed signals and simultaneous suppression of the noise effect. Filter banks that realize fast computational algorithms are synthesized. The wavelet approach is exemplified by solution of deconvolution, decorrelation, and differentiation problems.