Performance comparison of time-frequency distributions for ultrasonic nondestructive testing

In the nondestructive testing of materials ultrasonic backscattered echoes often exhibit critical time and frequency information. The time-frequency (t-f) analysis of ultrasonic signals instantaneously reveals the frequency and time of arrival of target echoes which help to characterize the target. But varying results are obtained by applying different t-f algorithms to ultrasonic signals. Performance evaluation of t-f algorithms for ultrasonic testing is useful for the choice of an algorithm. In this paper we present a performance comparison of t-f algorithms for ultrasonic applications. In particular, we compare the t-f algorithms based on the detection of the time of arrival and frequency of target echoes, estimation of instantaneous frequency of multiple target echoes, and cross-terms attenuation and concentration. These characteristics are studied by applying Wigner-Ville distribution (WVD), exponential distribution (ED), Gabor transform (GT), and wavelet transform (WT) to simulated and experimental ultrasonic data.