The University of Southampton
University of Southampton Institutional Repository

Signal reconstruction from partial or modified linear time-frequency representations

Signal reconstruction from partial or modified linear time-frequency representations
Signal reconstruction from partial or modified linear time-frequency representations

The first section of this thesis reviews such signal dependent techniques and describes a method of inversion for the Reassigned Spectrogram through use of its link to the phase of the Short Time Fourier Transform (ST-FT).

In order to achieve this inversion, signal reconstruction is required using only knowledge of the phase of the ST-FT.  Such an inversion is shown to be possible to  within an overall amplitude constant given mild conditions upon the degree of overlap used in the computation of the ST-FT.  Complimentary to signal reconstruction from knowledge of the phase of the ST-FT, is reconstruction from its magnitude.  Two distinct approaches are described.  The first previously given in the literature, iteratively applies the Minimum Least Squares (MLS) inversion of the complete (amplitude and phase) ST-FT.  The second uses an analogous approach to that used for signal reconstruction from phase.  Although the first approach proves more robust to non-valid input TF, both of these techniques require a similar degree of overlap to the phase case, and both reconstruct the signal to within an overall phase constant.  The description of the theory concludes with a discussion of the set of Generalised Wavelet Transforms (GWTs), of which both the SF-FT and WT are members.  After defining the set of GWTs, descriptions are given for the MLS inversion for complete GWT information, and for signal reconstruction from either phase or amplitude.

The thesis concludes by using the MLS based technique to create signals from modified or synthetic spectrograms generated using heart sounds.  The first application of this is to extend the duration of heart sounds.  Temporal extension in this fashion has the ability to extend the signals in time without affecting the spectrum of the sound.  In addition, it does not require the use of a model as in matching pursuit based methods described in the literature.  The second application is to create at time-series from a synthetic spectrogram constructed by averaging the spectrograms of a patient’s heart murmur.  Such averaging cannot take place in the time-domain owing to the random nature of the flow noise in a murmur.

University of Southampton
Lopes, David Manuel Baptista
Lopes, David Manuel Baptista

Lopes, David Manuel Baptista (2000) Signal reconstruction from partial or modified linear time-frequency representations. University of Southampton, Doctoral Thesis.

Record type: Thesis (Doctoral)

Abstract

The first section of this thesis reviews such signal dependent techniques and describes a method of inversion for the Reassigned Spectrogram through use of its link to the phase of the Short Time Fourier Transform (ST-FT).

In order to achieve this inversion, signal reconstruction is required using only knowledge of the phase of the ST-FT.  Such an inversion is shown to be possible to  within an overall amplitude constant given mild conditions upon the degree of overlap used in the computation of the ST-FT.  Complimentary to signal reconstruction from knowledge of the phase of the ST-FT, is reconstruction from its magnitude.  Two distinct approaches are described.  The first previously given in the literature, iteratively applies the Minimum Least Squares (MLS) inversion of the complete (amplitude and phase) ST-FT.  The second uses an analogous approach to that used for signal reconstruction from phase.  Although the first approach proves more robust to non-valid input TF, both of these techniques require a similar degree of overlap to the phase case, and both reconstruct the signal to within an overall phase constant.  The description of the theory concludes with a discussion of the set of Generalised Wavelet Transforms (GWTs), of which both the SF-FT and WT are members.  After defining the set of GWTs, descriptions are given for the MLS inversion for complete GWT information, and for signal reconstruction from either phase or amplitude.

The thesis concludes by using the MLS based technique to create signals from modified or synthetic spectrograms generated using heart sounds.  The first application of this is to extend the duration of heart sounds.  Temporal extension in this fashion has the ability to extend the signals in time without affecting the spectrum of the sound.  In addition, it does not require the use of a model as in matching pursuit based methods described in the literature.  The second application is to create at time-series from a synthetic spectrogram constructed by averaging the spectrograms of a patient’s heart murmur.  Such averaging cannot take place in the time-domain owing to the random nature of the flow noise in a murmur.

Text
789670.pdf - Version of Record
Available under License University of Southampton Thesis Licence.
Download (24MB)
Archive
789670_data.zip - Version of Record
Available under License University of Southampton Thesis Licence.
Download (422kB)

More information

Published date: 2000

Identifiers

Local EPrints ID: 467061
URI: http://eprints.soton.ac.uk/id/eprint/467061
PURE UUID: a76b1bd0-9409-4ad2-90a4-4a5d80a130fd

Catalogue record

Date deposited: 05 Jul 2022 08:10
Last modified: 05 Jul 2022 08:10

Export record

Contributors

Author: David Manuel Baptista Lopes

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×