The Pre-Processing of Ultrasonic Array Data for Time-Reversal-Based Imaging Algorithm

Chengguang Fan,Lei Yang,Yong Zhao

DOI: https://doi.org/10.1109/fendt.2017.8584591

2017-01-01

Abstract:Ultrasonic time-reversal-based imaging algorithm is a typical super resolution imaging method for detecting and imaging point-like defects embedded in a homogeneous medium in the field of Nondestructive evaluation (NDE). In this paper, a pre-processing method for extracting the valid scattered signals, reflected by the defects in the test structure, from pulse-echo time traces under the noisy case is proposed. A sub-area containing the defects is approximately assessed firstly via the conventional beamforming imaging algorithm, and then the gating point of window function is calculated, which can be applied to extract the valid scattered signals even when they are totally covered by the background noise. This proposed method is tested with experimental ultrasonic array data collected using the full matrix capture process. The experimental result implemented on a block of copper with two side-drilled holes (SDHs) shows that the proposed pre-processing method can extract the scattered signals from the pulse-echo time traces. Then multi-frequency time-reversal with multiple signal classification (MF-TR-MUSIC) over a given frequency range is applied to obtain the ultrasonic image of copper sample. According to the imaging result, the two SDHs can be located successfully.