February 7, 2012, 13.00, room K2.05
Kinect: where computer graphics meet computer vision and natural user interfaces
Georgios A. Triantafyllidis
During the last year there is a great research interest in applications based on consumer depth cameras (eg MS Kinect). Using these cheap consumer depth cameras, technologies that had been developed for expensive military or industrial purposes can now suddenly be used for creative and effective civilian purposes: building gestural interfaces to drive computer graphics, building cheap 3D scanners to produce point clouds and 3d meshes, using motion capture for easy 3D character animation, developing simple robotic navigation systems, etc.
In this seminar, first we will review the technology behind the consumer depth devices like Kinect and the available application programming interfaces (API). Next, a brief overview of the Kinect hacking scene will be presented. We will use the most popular API, which is the OpenNI/NITE, produced by the PrimeSense, the creator of the Kinect’s hardware and try to create graphics applications with Quartz Composer and Unity. We will then focus on application development by using the Kinect with Processing. Processing is an open source programming language and environment that uses the Java programming language graphics and other rich forms of media. The SimpleOpenNI library for Processing will be mainly used which supports many of OpenNI’s more advanced features and provides access to all of the Kinect data as well as a series of tools.
Georgios A. Triantafyllidis is an assistant professor at the Applied Informatics and Multimedia Dept of the Technological Educational Institute of Crete, Greece.