User-aware Content Orientation on Interactive Tabletop Surfaces

An increasing number of human computer interaction systems are employing interactive table surfaces. For these horizontally aligned screens, the orientation of text passages and any other 'oriented' graphical content is a common problem. A user will not be able to easily read the same text from different sides of such a table unless it adapts to his position. To overcome this problem, we present an interactive system that extends the interaction space from measuring the direct manipulation on the interaction plane to observing the user in the space above the table. Hence, the content of the graphical user interface can automatically be aligned to the position of the active user, which enables the ergonomic reading of a text. We present a viewpoint tracking system, which utilizes the Microsoft Kinect depth sensor accessed with the OpenNI framework. This system does not need initial pose calibration and smoothens the vision based tracking data. In a next step, we show the benefit of extending the interaction space for a drawing application that allows multiple users to work on automatically oriented, digital notepads while still being able to freely move around the table.