Automated photogrammetric close-range imaging of small invertebrates using acoustic levitation

Abstract Photogrammetry is gaining popularity in biodiversity research as it provides multiple views of the same specimen. However, handling the small specimens often requires fixing them with pins to position them in front of a camera, which can damage them. Acoustic levitation offers a solution by enabling lightweight objects to be suspended in air without physical contact. We demonstrate how small insects can be imaged from different perspectives using acoustic levitation to position and rotate them in mid-air. As opposed to other levitation systems with rotation control, prior knowledge of the geometry of the levitated object is not required. This way, automated multi-view imaging of insects with various morphologies is possible. To show the images are suitable to create a 3D representation, an image set is reconstructed in the 3D space using neural radiance fields (NeRF).