A hand gesture control framework on smart glasses

In this paper, we proposed a hand gesture control framework on smart glasses. Three different camera structures were presented to detect the hand portion, and the Moore’s Neighbor tracing algorithm detects the hand contour more efficiently and automatically. We not only refined the skin-color model but also improved the Chamfer matching method for the robust and effective gesture recognition.

A demonstration has been implemented by using the hand gesture control framework. Several gestures are pre-defined for various functions, such as selecting a virtual 3D object, rotating, zooming in or zooming out, and changing display properties of the 3D object.