The planned project is to employ safe, noninvasive motion capture and 3D visualization tools to study canine puppy socialization behaviors in a newly developed portable motion capture unit. While individual interactions have been widely described in the veterinary literature, few recorded observations of group dynamics are available. Using an array of synchronized structured light sensors to capture a collection of moving objects is a feasible technique for recording and allowing the quantification of puppy social interactions. Structured light sensors, such as the Microsoft Kinect, can capture the depth of a scene over time. This technique for capturing depth is faster than a laser scanner because it covers the entire field every frame and is more suited toward moving objects. Several of these sensors can be combined together to provide a more complete image of a 3D scene of objects. With the array of sensors calibrated to the same environment, their depths can be combined to produce an aggregate point cloud. The point cloud scene can be segmented to uniquely identify and track separate objects. These data can be used in future studies with the use of machine learning algorithms to computationally identify social behaviors.
Thomas Tucker, Creative Technology, School of Visual Arts, College of Architecture andUrban Studies at Virginia Tech
Bess J. Pierce, Department of Small Animal Clinical Sciences, Virginia-Maryland Regional College of Veterinary Medicine at Virginia Tech
Jeryl C. Jones, Division of Animal and Nutritional Sciences, Davis College at West Virginia University
Matthew Swarts, Digital Building Lab, SimTigrate Design Lab at Georgia Tech
Architecture & Urban Studies