A comparison of wavelet deconvolution techniques for ultrasonic NDT

Abstract

Wavelet deconvolution is a problem of fundamental importance in many signal processing applications. The authors are concerned with deconvolution to extract impulse responses of materials tested in ultrasonic nondestructive testing (NDT). The impulse responses can provide vital information about the flaws and material properties. Several deconvolution techniques including Wiener filtering in three versions, the spiking filter, time domain deconvolution, L1 norm deconvolution, and others are examined and critical comparisons are made for the ultrasonic NDT data.<>