Title: An underwater acoustic imaging method for pattern recognition

Author(s): HE Xianzhong; XU Feng;

Affiliation(s): Institute of Acoustics, Chinese Academy of Sciences; University of Chinese Academy of Sciences

Abstract: An underwater acoustic imaging method for pattern recognition in a free-field environment was proposed, where angular domain filtering is used to separate plane waves from different directions of space in the array element domain. In particular, a series of angular domain bandpass filters are designed, where the output from these filters is the snapshot of data from an incident plane wave within the angular passband in the array element domain. Our approach is potentially applicable to pattern recognition algorithms, such as target recognition and bottom classification, which need the characteristic data of the array element domain. Moreover, underwater acoustic imaging may be performed by using the amplitude of snapshots from the angular domain filters as the pixel values. The angular domain filtering method may be used for underwater acoustic imaging to obtain the desired sound image, and each pixel corresponds to a snapshot of data generated by far-field plane wave coming from the target and hitting the receiving array. Simulation and experimental results show that the sound images obtained by filtering in the angular domain are close to those obtained by the conventional beamforming method. We also find that the two imaging methods are robust to the transducer array element errors.