The Microsoft Kinect has recently grown to prominence as a widely used 3D sensor for both academics and hobbyists alike. This thesis presents a Kinect-oriented framework for reproducing physical objects using open or closed source software, GPU-accelerated hardware, and a 3D printer. Specifics of data capture, surface reconstruction, and manufacture are discussed, along with examples of reproducing mechanical and biological models. Finally, future systematic improvements and additional application areas are discussed.