• Heino Jørgensen
4. term, Medialogy, Master (Master Programme)
This project focuses on using natural gestures to control a slideshow presentation. Using the Microsoft Kinect an approach is given for tracking the users hand and fingers, in order to supply for a gesture based interaction with the presentation. The method proposed for finger tracking uses skin color detection combined with feature extraction of fingertips, implemented using OpenCV in combination with OpenKinect. For the final experiment though, an application only utilizing hand tracking was implemented, due to the instability of the proposed finger tracking. The final application is implemented in Unity, using the NITE library from OpenNI. It allows a presenter to move through a presentation by swiping as well as highlighting text and objects on the slides. The final experiment tested the audience perception of the gesture based presentation, compared to a traditional presentation using mouse and keyboard. The test showed that while the gesture based presentation was neither better nor worse for conveying information to the audience, it seemed to distract the audience more, because of the presenter’s movement.
Publication date27 May 2011
Number of pages72
ID: 52594037