The use of instantaneous phase for improving sparse arrays images

Abstract

Sparse arrays have pitch larger than half-wavelength (lambda/2) and there is a reduced number of elements in comparison with a full-populated array. Consequently, there is a reduction in cost, data acquisition and processing. However, conventional beamforming techniques result in large side and grating lobes, and consequently in image artifacts. In this work the instantaneous phase of the signals is used in a beamforming technique instead of the instantaneous amplitudes to improve images obtained from sparse arrays configurations. A threshold based on a statistical analysis and the number of signals used for imaging is applied to each pixel, in order to determine if that pixel is related to a defect or not. Three sets of data are used to evaluate the technique, considering medical and non-destructive testing: a simulated point spread function, a medical phantom and an aluminum plate with 2 lambda-, 7 lambda- and lambda-pitch, respectively. The conventional amplitude image is superposed by the image improved by the instantaneous phase, increasing the reflectors detectability and reducing artifacts for all cases, as well as dead zone for the tested plate.