A team of researchers from Arizona State University (ASU) has unveiled a new solution that allows the precise tracking of finger movement for use with augmented and virtual reality (AR/VR) headsets.
The solution — called FMKit — aims to solve the problem of input methods (how to interact with objects and UIs) when using AR or VR, with most methods currently relying on a handheld controller of some sort.
With FMKit, AR and VR headsets can accurately track individual finger motions and recognize in-air handwriting, allowing for new forms of secure user authentication.
Fingertip writing could act as the unique signature used to authenticate individuals, or the system could identify users based on passwords written using their fingers. Both of these methods come as a potentially more secure alternative to using a handheld controller to type or select words for authentication.
FMKit differs from other hand tracking solutions such as Leap Motion in that it can not only map and record an individual finger’s path through three-dimensional space but compare it against four data sets of other handwriting samples.
In its current state, FMKit supports either a Leap Motion ‘controller’ (which works at 110 scans per second), or a custom inertia-measuring data glove (50 scans per second). The Leap Motion input method achieves more than 93 percent accuracy while the glove is slightly better at 96 percent accuracy.