Combined use of color and depth information for hand gesture recognition

In the last years several human machine interfaces has been developed mainly in the gaming industry. The work of this thesis with reference to the Microsoft Kinect device, which provide to get substantially a color and a depth images, proposes an algorithm to track the positions of the hands in a scene for hand gesture recognition applications. Depth image coordinates are converted to the image color reference system using the Heikkil ̈a calibration model