A Non-linear Mapping Technique for Bare-hand Interaction in Large Virtual Environments

Computer vision-based bare-hand tracking systems have become popular as they free the users from wearing any hardware devices (e.g. gloves). However, due to the limitations in tracking methods and human body constraints, interaction methods developed for other 3D interaction devices can have problems when applied to computer vision-based bare-hand tracking systems. Therefore, there is a need to develop interaction techniques suitable for bare-hand input. In this paper we investigate a non-linear mapping technique for bare-hand input in large-scale virtual environments based on physiological constraints. This method not only enables high precision interaction with objects at both near and far distances, but also allows users to manipulate objects efficiently with significant translation and rotation. Through a user study, we show that the proposed method can be fairly effective, precise, more usable, and less stressful compared to two more traditional baseline interaction techniques