Hand Motion Gesture Recognition System

System Diagram


Hand motion is an effective means of human communications in real world. It might be also true in virtual environment but the use of hand motion is still limited to a simple application. This project is to develop a generic software package for hand motion recognition using hidden markov models, with which user interface designers will be able to build a multimodal input system. Hand motion recongnition package provides functions for converting spacial hand motions into symbolic forms, that makes it easy to combine with other inputs such as speech recogniton, static hand postures, etc.

Besides sign language recognition, hand motion will be able to be used for specifying shapes and sizes in CAD/CAM application, enhancing degree of freedom for hand posture command inputs and menu selections, and recognizing handwritten spatial charactors.

This project involves the development of C++ software packages that provides various functions for handling hand motions. Currently it consists of three modules: Hidden Markov Models module, Plane fitting module and device interface modules. The device interface module is a enhancement of
GloveGRASP software package that was developed earlier.

Related works on gesture recognition

$HITLab: HMRS.html,v 1.2 2001/01/23 17:42:45 perseant Exp $