This dissertation introduces a new idea for frequency
domain deconvolution of speech signals. The idea is to
approximate a unit sample by a Gaussian pulse. The
deconvolution is performed by designing a glottal inverse
filter that will estimate the excitation of the vocal tract.
Then, the excitation will be used...
More efficient methods are needed for the in-situ evaluation of mass transfer parameters which describe the movement of solutes through aquifer material. The objective of this study was to develop a method for estimating diffusion rate and capacity coefficients using a single-well, "push-pull" tracer test. The method consists of the...
Frequency domain design of deconvolution filters is
studied using Fast Fourier Transform. Filters for both the
spike and the Gaussian reflector wavelet are obtained. True
deconvolution filters are infinitely long IIR filters,
and frequency domain analysis is an effective way of
finding its optimum finite length approximation for an
arbitary...
The resolution of multiple overlapped reflections is
achieved by designing inverse filters using frequency
domain deconvolution. A careful study reveals a number of
problems involving frequency domain design of spike inverse
filters. These include the presence of noise and aliasing
in both the filter response and its output resulting from...