Real time hand tracking and 3d gesture recognition for interactive interfaces using hmm



In this study we have developed a humancomputer interaction interface (HCI) based on real time hand tracking and 3D dynamic gesture recognition using Hidden Markov Models (HMM). We propose a system, which captures and recognizes hand gestures of the user wearing a colored glove, where the hand coordinates are obtained via 3D reconstruction from stereo. In recognition of eight defined gestures, the system was able to attain 98.75% accuracy in 160 trials, using a pair of ordinary web cams. We provide supplementary features such as an interactive training system and a self-calibration utility for the cameras. The overall model is designed to be a simple and robust gestural interface prototype for various PC applications.


Cem Keskin
Ayse Naz Erkan
Lale Akarun