Ultrasonic Signal Processing Based on the Combined Use of Empirical Mode Decomposition and Split Spectrum Processing Using the Prism Technique

Abstract

In this work we present the application of Empirical Mode Decomposition (EMD) and Split Spectrum Processing (SSP) in non-destructive testing of materials using ultrasound. The SSP as well as the EMD technique enable to enhance target detection and the visibility of reflected echoes. The use of the EMD technique before using SSP makes the detection of any desired target-echo more flexible, when looking for a good solution to the coherent noise problem. Some types of coherent noise can be removed by frequency filtering but only provided that the amplitude spectrum of the noise does not overlap the spectrum of the desired signal. In the most practical situations, this noise may comprise refractions, multiple reflections, reflected refractions or refracted reflections, and coherent noise from grains and scatters with different sizes. The signals were obtained using a technique applied in pulse-echo mode, known as the prism technique and processed in Matlab environment. Tests carried out on trapezoidal-prism shaped specimens of mortar highlight the capability of this signal processing technique.