Hand Gesture Recognition

Controlling the computer through hand movement, without touching a device

Hand gesture driven applications :

Imagine the human-computer interaction of the future: Imagine a 3D-application where you can move and rotate objects simply by moving and rotating your hand - all without touching any input device.

You can get more than 6 degrees of freedom simultanously (X-, Y-, Z-rotation, X-, Y-, Z-translation, and additional degrees calculatable from the handform, e.g. how strong the thumb is spreaded from the hand). The user gets the impression of an easy and intuitive interface, because he can change tools or give commands by showing another hand gesture.
For example: Moving and rotating is enabled when the hand is open, selecting is enabled when the index finger is shown, zooming when both the index finger and the thumb are shown.
A lot of dialog states can be saved in this way.

Hand Gesture Recognition in detail:

Your hand is recorded by a video camera and digitized by special hardware. An image recognition software uses this picture data to constantly determine the current position and form ("gesture") of your hand. This information is then used to drive an application.

For control purposes, the video, the hand contour with center of gravity and inertia axis, the signature calculated from the contour and the currently gesture parameters are displayed on the screen.

The image recognition software that extracts the gesture information is called OGR (Optical Gesture Recognition). I wrote OGR as part of my Master's Thesis. OGR bases on wxWindows, a portable C++ class library, so it is portable to other platforms (currently running on DEC-UNIX, Motif) and mostly portable to other class libraries. OGR can be used as class library for other applications !

So What ?

Controlling computers via hand gestures can make many applications more intuitive than using keyboard or even mouse input. Since desktop cameras are becoming standard workplace equipment (e.g., for video conferencing), using those cameras for gesture recognition does not require any additional hardware. Finally, public information kiosks could use this technology to solve hygienic and input device wear problems.

Want to see screenshots ?

Here's a screenshot (4K) of the main window of the gesture server.

Look at this screenshot (15K) of the print preview dialog of the gesture server.

You want some files ?

Note: All these files are provided as information and for testing. I am the owner of the provided source code, binaries, and documents. They are not public domain!!! Private use and use for testing is granted. For commercial use, please cantact me.

Back