Interaction Spaces: Interactive Spatial Areas to Control Smart Environments

Abstract

Throughout recent years new input modalities have found their way into consumer electronics. Recognizing body posture and gestures in the three dimensional space is now possible using hardware that is available for about 100 EUR. We aim at providing a system to convert any environment into an interactive space. Hence, we created a system that is able to detect the user s body in three dimensions and to determine the presence of body parts at pre-defined/user-defined locations in order to trigger actions of the environment. We built a first Kinect-based prototype where users can define trigger areas and link them to suitable actions. We then conducted a study to evaluate the usability of the system and how size and memorability of spaces affect user performance with regard to trigger area tasks. Results show that it is easy to remember trigger area positions. With increasing area size the task completion time goes down while error rates go up.