Multiple-Camera Based Hand Pose Estimation using Skeleton Images

Abstract

We describe a method to detect hand position, posture and finger bendings using multiple camera images. Stable detection of position and posture can be performed by using skeleton images. We confirmed the stability of those through experiments. This system can be used as a user interface device in a virtual environment, replacing glove type devices and overcoming most disadvantages of contact type devices. Future work includes development of parameter adjustment mechanism for hand movement and verification of availability as a gesture based man-machine interface system.