Preprocessing of the full matrix capture data for time-reversal-based super-resolution imaging

In this paper, a preprocessing method based on the plane B-scan, the typical beamforming imaging algorithm, is proposed to extract the valid scattered signals reflected by defects from pulse-echo time traces in a case involving noisy materials. The method is tested with experimental ultrasonic array data collected using the full matrix capture (FMC) process and the experiment is implemented on a block of copper with two side-drilled holes (SDHs). It is shown that the proposed preprocessing method can successfully extract the scattered signals, even when they are totally covered by the background noise, via approximately assessing the position of a defect and calculating the gating point of the window function by selecting a subregion. Then, the scattered signals are post-processed by multi-frequency time reversal with multiple signal classification (MF-TR-MUSIC) over a given frequency bandwidth to yield an image with improved resolution. According to the MF-TR-MUSIC imaging result, the two SDHs can be located. In addition, the influence of the selected subregion on imaging performance is also investigated.