An automated visual tracking measurement for quantifying wing and body motion of free-flying houseflies

Abstract

Pioneering discoveries revealed that flying insects actively regulate body appendages such as wings, legs and abdomen to stay aloft. However, the initial stage of capturing their motion during flight is rather challenging and time-consuming, especially during the digitization of lengthy video images. Therefore, our development of an automated visual tracking system will greatly provide a full access to insect's body and wing dynamics during flight. By using the positional dataset obtained from the digitized images which earlier captured by an automated time-resolved high-speed videography, we thus further three-dimensionally reconstructed body and wing dynamics of housefly Musca domestica. We validated and further compared the automated digitization with manual tracking. Our analysis estimates that motions along z-axis yields higher differences (16 ± 28.19 µm for thorax and 13 ± 99.19 µm for wingtip) because it orthogonally points to the cameras, which lead to acceptable inaccuracies of calibration coefficients due to the limited depth of focus.