Palm area detection for reliable hand gesture recognition
Giulio Marin, M. Fraccaro, Mauro Donadeo, Fabio Dominio, Pietro Zanuttigh · 2013
Abstract—Hand gesture recognition applications require as a first step a reliable identification of the hand region and its subdivision into fingers and palm areas. In this work the hand is firstly extracted from the depth data acquired by a depth sensor (e.g., Microsoft Kinect). Then the center of the palm and the hand orientation are identified. Finally circular and elliptical shapes are fitted on the extracted samples in order to reliably identify the palm and fingers area. The proposed approach has been tested inside a simple gesture recognition scheme and preliminary results show its reliability. I.