Location
Brunswick, ME
Start Date
1-1-1992 12:00 AM
Description
The application of signal processing techniques to nondestructive evaluation(NDE) has proven successful for years. The research on this issue has been directed into two types of techniques, namely, the time-domain methods and frequency methods. Techniques of frequency-domain analysis such as Fourier transform and power spectral density reveal important information concerning decomposition of frequency components which may not be available from the time-domain analysis. However, in the case of ultrasonic NDE, the observed echo from a material can be viewed as a superposition of successive time arrivals of different components each characterized by different spectral contents. Both the time domain waveform and the spectrum can be very confusing and are simultaneously relevant for a complete description of a signal [2]. Neither the spectral analysis nor the time-domain methods alone can provide needed information. Therefore, mixed tools are often necessary to be introduced to overcome the limits imposed on these two mutually complementary families of techniques. These tools are generally referred to as time-frequency distribution or time-frequency analysis and have been found useful in numerous applications such as speech processing, sonar analysis, detection of electrocardiography (EEG) signals and X-ray diffraction. Among them, the Wigner distribution has been studied intensively in the last few years due to its high resolution property
Book Title
Review of Progress in Quantitative Nondestructive Evaluation
Volume
11A
Chapter
Chapter 3: Interpretive Signal Processing and Image Reconstruction
Section
Signal Processing
Pages
967-974
DOI
10.1007/978-1-4615-3344-3_124
Copyright Owner
Springer-Verlag US
Copyright Date
January 1992
Language
en
File Format
application/pdf
On the Use of Wigner Distribution in Ultrasonic NDE
Brunswick, ME
The application of signal processing techniques to nondestructive evaluation(NDE) has proven successful for years. The research on this issue has been directed into two types of techniques, namely, the time-domain methods and frequency methods. Techniques of frequency-domain analysis such as Fourier transform and power spectral density reveal important information concerning decomposition of frequency components which may not be available from the time-domain analysis. However, in the case of ultrasonic NDE, the observed echo from a material can be viewed as a superposition of successive time arrivals of different components each characterized by different spectral contents. Both the time domain waveform and the spectrum can be very confusing and are simultaneously relevant for a complete description of a signal [2]. Neither the spectral analysis nor the time-domain methods alone can provide needed information. Therefore, mixed tools are often necessary to be introduced to overcome the limits imposed on these two mutually complementary families of techniques. These tools are generally referred to as time-frequency distribution or time-frequency analysis and have been found useful in numerous applications such as speech processing, sonar analysis, detection of electrocardiography (EEG) signals and X-ray diffraction. Among them, the Wigner distribution has been studied intensively in the last few years due to its high resolution property