Real-Time and High-Precision Acoustic Holographic Field Reconstruction Based on Iterative Unsupervised Learning

Abstract The ability to shape a specific acoustic field is crucial for a variety of applications, especially in contactless operation, where it shows great advantages. Acoustic holography reconstructs the desired acoustic field by encoding a three-dimensional (3D) acoustic field into a two-dimensional (2D) acoustic hologram. Most conventional acoustic holography, such as 3D-printed acoustic lenses and phased arrays of transducers (PAT), has limitations in terms of dynamics and phase fidelity. To address these issues, in this article, a method is proposed that combines the high-fidelity benefits of acoustic holograms with the dynamic control of phased arrays in the ultrasonic frequency range. An iterative unsupervised learning algorithm is also proposed to compute the desired source phase holograms and move the position of the projected holograms through the phase delay to control the output acoustic waves and obtain the desired acoustic field. Experiments show that the method outperforms previous algorithms in terms of phase fidelity and real-time performance, demonstrating the future potential of acoustic holography in the field of automated noncontact particle manipulation.