Application of Auto-tracking to the Study of Insect Body Kinematics in Maneuver Flight

Shreyas Subramanian

Abstract: There is a need to explain the complex phenomena that underlies the seemingly effortless flight modes of the dragonfly (Infra-order Anisoptera). However, measuring the body kinematics during flight is labor intensive. Thus a robust system was developed that automatically tracks and quantifies the body kinematics of a dragonfly during voluntary and escape take-offs, as well as maneuvers. Ultimately, the tool, which was developed using a custom code in C++ using the open source library OpenCV (Open Computer Vision), would be used to analyze bulk samples of high speed videos providing raw images at the rate of approximately 1000 frames per second from pair-wise orthogonal positions in space. As a result, there would be a considerably large database of information which may then be used to formulate, generalize and classify standard flight strategies used. Perceptibly, there is also a need to validate the outputs of this tool by comparing it to the outputs of a manual reconstruction.