A Nautilus bionic multi-information fusion compressed-sensing acoustic imaging device

Abstract

Sound waves carry abundant physical information essential for environmental perception. Traditional sensor-array-based sound-source localization methods suffer from drawbacks such as large system size and complex data processing. Existing compressive-sensing imaging methods can realize sound identification, but the reliance on highly anisotropic metamaterials makes it difficult for them to achieve high-precision sound-source localization with relatively regular low-loss devices. Inspired by the Nautilus structure, we propose a bionic metamaterial multi-information fusion compressed-sensing acoustic imaging device for sound localization and identification. By imitating the spiral geometry of the Nautilus, the regular metamaterial design strategy reduces the structural complexity and the sound loss. We introduce a multi-information fusion method to decrease anisotropic reliance and enhance compressed-sensing acoustic imaging capabilities. The proposed positioning device can identify multiple broadband sound sources with a high identification success rate even in noisy environments, which shows wide application prospects in medical inspection and human-computer interaction.