An efficient wavenumber algorithm towards real-time ultrasonic full-matrix imaging of multi-layered medium

Ultrasonic full-matrix imaging can sense small defects in non-destructive evaluation using full-matrix capture (FMC). However, the accompanying heavy computational burden limits FMC's applications, especially for real-time scenarios. Currently, the full-matrix imaging for homogeneous medium has progressively advanced, but these algorithms become inaccu-rate and inefficient for the multi-layered medium. To overcome this problem, an efficient wavenumber algorithm for ultrasonic full-matrix imaging of multi-layered medium is proposed in this paper. The proposed method utilises wavefield extrapolation to migrate the transmitting and receiving transducers virtually to the interfaces between different layers. It then reconstructs the image of each layer with Stolt interpolation. The proposed method is developed based on a wave model, and thus is more rigorous than the ray-based theory in mathematics. Moreover, the calculation in the wavenumber domain dramatically reduces the computational cost. In comparison with well-established root-mean-square based total focusing method and extended phase shift migration, the proposed method performed better in terms of imaging quality and efficiency in the simulations and experiments. Specifically, it achieves 6-10 Hz imaging frame rates for ultrasonic full-matrix imaging of multi-layered medium using a 64-element phased array, and may find various real-time imaging applications in practice. (c) 2020 Published by Elsevier Ltd.