Ultrasonic array time-reversal based super resolution imaging

chengguang fan,mengchun pan,dixiang chen,yunze he,feilu luo

DOI: https://doi.org/10.1109/FENDT.2014.6928235

2014-01-01

Abstract:In this paper, time-reversal with multiple signal classification (termed TR-MUSIC), the widely used time-reversal based imaging algorithm, has been introduced into the area of ultrasonic NDE. This method can break diffraction limit which is given by Rayleigh criterion and achieve super resolution imaging. Here, resolution means the ability of an imaging method to resolve closely spaced targets. A parameter named array performance indictor (API) is utilized to quantitatively assess the TR-MUSIC and beamforming imaging method. It is shown that for the situation that the size of defect is smaller than the wavelength, TR-MUSIC can successfully distinguish poorly-resolved targets that cannot be resolved by traditional beamforming imaging algorithm, and a resolution which is about 2.2 times that calculated by Rayleigh criterion can be obtained using TR-MUSIC. In addition, the super-resolved ability of TR-MUSIC is stable for different noise level material and the API obtained from TR-MUSIC is smaller than that of beamforming method, which has been experimentally demonstrated.