Publications
Volumetric Video Motion Detection for Unobtrusive Human-Computer Interaction
The computer vision field has undergone a revolution of sorts in the past five years. Moore's law has driven real-time image processing from the domain of dedicated, expensive hardware, to the domain of commercial off-the-shelf computers. This thesis describes their work on the design, analysis and implementation of a Real-Time Shape from Silhouette Sensor (RT S{sup 3}). The system produces time-varying volumetric data at real-time rates (10-30Hz). The data is in the form of binary volumetric images. Until recently, using this technique in a real-time system was impractical due to the computational burden. In this thesis they review the previous work in the field, and derive the mathematics behind volumetric calibration, silhouette extraction, and shape-from-silhouette. For the sensor implementation, they use four color camera/framegrabber pairs and a single high-end Pentium III computer. The color cameras were configured to observe a common volume. This hardware uses the RT S{sup 3} software to track volumetric motion. Two types of shape-from-silhouette algorithms were implemented and their relative performance was compared. They have also explored an application of this sensor to markerless motion tracking. In his recent review of work done in motion tracking Gavrila states that results of markerless vision based 3D tracking are still limited. The method proposed in this paper not only expands upon the previous work but will also attempt to overcome these limitations.