Location

Seattle, WA

Start Date

1-1-1996 12:00 AM

Description

Wavelet transform has already been shown as a useful tool for the interpretation and the enhancement of ultrasonic data in the context of nondestructive evaluation [1–3]. Main applications of the wavelet transform are signal analysis in the time-frequency domain, data compression and now signal processing. Comparisons with other time-frequency representations like short time Fourier transform [1] and Wigner-ville transform [2] have shown the usefulness of the continuous wavelet transform for signal analysis: this method is well adapted to localize in time both high and low frequencies and does not introduce interference terms. Another important property is that signal reconstruction can be achieved from wavelet decomposition. This ability allows one to do signal processing in the time-frequency plane. Earlier work has shown the possibilities to use the wavelet transform as a filter for signal-to-noise ratio enhancement [2] by reconstructing the signal after applying energy thresholding in the time-frequency domain. This reconstruction does not involve global averaging in time or frequency domain because of the good localization of the wavelet coefficients in both domains.

Volume

15A

Chapter

Chapter 3: Signal Processing and Image Analysis

Section

Signal Processing

Pages

749-755

DOI

10.1007/978-1-4613-0383-1_98

Language

en

File Format

application/pdf

Share

COinS
 
Jan 1st, 12:00 AM

Echo Extraction from an Ultrasonic Signal Using Continuous Wavelet Transform

Seattle, WA

Wavelet transform has already been shown as a useful tool for the interpretation and the enhancement of ultrasonic data in the context of nondestructive evaluation [1–3]. Main applications of the wavelet transform are signal analysis in the time-frequency domain, data compression and now signal processing. Comparisons with other time-frequency representations like short time Fourier transform [1] and Wigner-ville transform [2] have shown the usefulness of the continuous wavelet transform for signal analysis: this method is well adapted to localize in time both high and low frequencies and does not introduce interference terms. Another important property is that signal reconstruction can be achieved from wavelet decomposition. This ability allows one to do signal processing in the time-frequency plane. Earlier work has shown the possibilities to use the wavelet transform as a filter for signal-to-noise ratio enhancement [2] by reconstructing the signal after applying energy thresholding in the time-frequency domain. This reconstruction does not involve global averaging in time or frequency domain because of the good localization of the wavelet coefficients in both domains.