The real-time generation of 3-D acoustic images by an active system deploying a planar array of transducers is difficult to implement because of the hardware cost and the associated computational burden. Recently, an exact method based on chirp zeta transform (CZT) beamforming has been proposed to dramatically reduce the computational burden of 3-D wideband imaging in the near- and far-field regions. However, the computational advantage of such a method decreases if the array is thinned. In this paper, the performance of CZT beamforming applied to sparse planar arrays is investigated and compared with those of other beamforming techniques.
To address the sparse array challenge, a processing technique, called pruned CZT beamforming, is proposed, which exhibits a computational burden lower than that of the previous CZT beamforming technique while providing a significant advantage over the direct implementation of the frequency-domain beamforming and is applicable when the planar array is radically thinned.