Characterising a High-Speed 3D Surface Scanner for Bat Behaviour Study

There has been an increasing interest in studying animal behaviours [1]. In the EU funded project “ChiRoPing: Developing Versatile and Robust perception using Sonar Systems that Integrate Active Sensing, Morphology and Behaviour” [2], the interest is placed on Chiroptera (bats) as bats are the nature’s expert in active sonar sensing [3]. Study of bats’ behaviour is not only of biological interest in the project but also helpful in engineering an active sonar sensing system similar to what bats possess.

One source of information to be investigated in this study is 3D morphology, and in particular how the morphologyof a bat species is related to their echolocation behaviours. To this end, dynamic 3D surface capture of bat heads in flight during echolocation is required. Together with sonar data collected by acoustic sensors, the dynamic 3D shape data can enable biologists and acoustic experts to examine the details of bats’ sonar system in vivo. A high-speed stereo photogrammetry based 3D scanner is employed to perform the 3D capture. The main reasons are twofold. First for capture speed, stereo photogrammetry only requires passive capture of images from two views, thereby allowing fast data recording; second for data consistency, stereo photogrammetry could generate effective establishment of 3D point correspondence between the captured frames allowing a consistent shape analysis over a entire 3D sequence. In this paper we report our scanner
characterisation experiments.