Compressed Sensing for Full Matrix Capture Data Based on Optimal Reconstruction Algorithm

Ultrasonic phased array total focusing method (TFM) attracts much interest because of high image resolution and large dynamic range. However, the data acquisition method relied on by TFM is full matrix capture (FMC), which generates huge data and brings great pressure. In this paper, the feasibility of compressed sensing (CS) in the reconstruction of ultrasonic FMC data is investigated. Firstly, four greedy algorithms are used to compress and reconstruct the ultrasonic signal, and the optimal algorithm is selected according to the reconstruction accuracy and the influence of sampling rate is considered. Secondly, the optimal algorithm is applied to the reconstruction of undersampled FMC data, and TFM is performed based on reconstructed FMC data. The simulation results demonstrate that excellent reconstruction can be obtained using measurement points less than the Nyquist sampling limitation required, the spatial resolution of TFM imaging of FMC data reconstructed with CS remained unchanged. Finally, experimental verification is carried out. Although the experimental results are not as ideal as the simulation results, FMC data can still be accurately recovered with a small number of measurements and the identification of defects by TFM is guaranteed.