University of Twente Human Media Interaction

Multi-touch tables are often used to display 3D imagery on the 2D surface of the table. Interacting with these 3D images through the 2D plane can be unintuitive and complicated. In short the goal of the research described in this thesis was to find out if interaction with 3D objects and environments on multi-touch tables could be made more intuitive by adding a third dimension to the input capabilities. A prototype was constructed using a Rear Diffusion Illumination based multi-touch table and the Flock of Birds tracking system. The 3D position and orientation of the user’s hand were tracked using the Flock of Birds system. Existing software was combined with software developed specifically for this project to allow a user to interact with applications such as Google Earth and Jmol. The limitations of the tracking sensor, the functionality of the applications and the goal of creating intuitive interaction were combined to design an individual interaction scheme for both applications (Google Earth and Jmol). For the evaluation of our prototype a user test was conducted. Our prototype was compared to a system that used purely 2D touch based input. The participants were asked to complete some tasks using the two systems and to fill out a questionnaire on their experiences. Existing questionnaires, on subjects such as usability and comfort, were used. The experiment was supposed to provide an answer to the question of which system performed better in terms of perceived usability, comfort, etc. Unfortunately the analysis of the responses to the questionnaires did not allow us to answer this question. The rather unexpected outcome of the experiment was that the participants always preferred the second system they used, independent of what it was (our prototype or the alternative system). The most plausible explanation of this result seems to be that the participants needed more time to get used to the tasks they were asked to complete. Introducing more practice time would most likely have resulted in more insight into which of the two systems the participants preferred. The comments and observations made during the experiment did indicate the participants’ interest in our prototype. This fact combined with the lessons learned from the research described in this thesis, with regards to both the prototype design and the experiment setup, warrant further research into this area of three dimensional interaction with multi-touch tables.