Reconstructing outside pass-band data to improve time resolution in ultrasonic detection

A combination deconvolution method is presented to increase the time resolution of an ultrasonic echo signal by reconstructing outside pass-band data. By the modified traditional Wiener inverse filter the filtered data are estimated, an autoregression (AR) model is obtained for these data inside the pass bands. Only the high frequency side data out of the pass bands are renewed by the autoregressive spectral extrapolation method. The optimized estimation is reconstructed using the combination of renewed and filtered data inside the pass bands. The results from the computer simulation and the experiments on measuring the thickness of thin multilayer materials show that this approach has improved time resolution performance.