Technical Report No. 221 - Abstract

K. A. Mohamed
Identifying Tabletop Corners with Angular- and Velocity-Type Trace-Features in Pen-Gestures

We present an effective corollary from our current works in pen-gesture technologies, originally in-tended for conventional digital screen environments, that gives significant influence when expanded into the domain of tabletop displays. By virtue of incorporating angular and velocity-type “trace-features” for pen-gesture cognition into current classical linear-classifiers, we are able to place a greater emphasis not on the similarity of the “visual representation” of the gestures, but rather, on determining exactly from which corners of the table the gestures were conceived. In other words, we can find out which one of the four users around the table is interfacing with the tabletop display at any particular time, and do this without the need for additional sensing paraphernalia. We find this especially useful when the hardware requirement is limited to the very basic, and when the digital tabletop displays offer only asynchronous input capabilities; i.e. only one anonymous digital pen input can be considered touching the screen correctly at any one instant. We illustrate this inception by discussing our treatment of “disoriented” pen-gestures with our version of a ‘manual’ Monopoly board game for the tabletop.