Tracking human gestures in three dimensions (3D), and using that information to meaningfully interact with computers, has become an enticing exercise. Bimber [1] provided these authors with the inspiration to improve on his algorithms and means of tracking wand movement. Such wand movement could be used to interact with certain computer games.

Bimber’s method included a six-degree of freedom (6DOF) means to concisely record data about change in location in 3D, as well as the change of orientation of the object being tracked. Sequential samples of these changes could be saved to represent particular meaningful motions. Sets of this type of data are designated to represent various gestures. Fuzzy logic is then applied, comparing the representative data to data collected from someone whose motions are being tracked. A close enough match is determined to be a meaningful gesture.

The alterations proposed by the authors included simplifying the type of data being collected (orientation only), changing Bimber’s weighting formula, and applying both of these changes at once. The number of experiments was somewhat limited, and the ability to gather sample data was sometimes affected by restrictions of the devices sensing visual motion. The most interesting result of this study is that standard data that was collected by one device (modeling a particular meaningful gesture) could be used successfully in identifying gestures tracked by a different device.

The tests cited in this paper are minimal, and conclusions are intended to spur more research. Portions of the paper will make sense to the average reader. Interpretation of the authors’ methodology, understanding the reasons behind the modification of data collection and its interpretation, and interpretation of the figure-based data are only feasible for those familiar with Bimber’s approach to recognizing gestures. This paper is probably only of interest to those planning a larger, related study.