Meeting Abstract
Birds fly effectively through complex windy environments, and in order to understand the strategies that enable them to do so, we need to determine the shape and movement of their wings. Previous studies show that even small perturbations in wing shape have dramatic aerodynamic effects, but these shape changes have not been quantified automatically at high temporal and spatial resolutions. Hence, we developed a custom 3D surface mapping method which uses high-speed cameras to view optimized grids of stripes projected onto a flying bird. By matching the stripes seen in each camera frame with the projected stripes, we can triangulate the 3D bird surface along the grid lines. Because the light is binary rather than grayscale, and each frame is separately analyzed, this method can capture rapidly moving objects at any frame rate, as long as the camera captures sufficient light contrast. In addition, the method is automated, non-invasive, and capable of measuring a shape volume by simultaneously reconstructing from multiple view angles. We use this technique to reconstruct the 3D surface of a parrotlet during flapping flight at 3200 fps. From this shape we can extract the airfoil, velocity, and angle of attack which allows us to analyze key dynamic parameters such as lift and drag. While this novel system is designed to quantify bird wing shape and motion, it is adaptable for tracking other quickly deforming objects including other animals, especially those which are difficult to reconstruct using other 3D tracking methods.
