On the fly 3D surface reconstruction: KinectFusion

Microsoft’s Kinect is rightfully getting a lot of attention from researchers. One snippet that caught my attention is a collaboration between Microsoft and a number of UK and Canada-based researchers. The result is KinectFusion.

Have a look for yourself:

The implications for virtual worlds are fairly obvious. The thing that particularly struck me is the dynamic capability of the approach even at this early stage – if something changes with the physical world environment, it is reflected virtually. For the education, science and health fields, to name three, this is huge.

One obvious example within my pet area of clinical simulation: a camera (with consent) is placed in a busy emergency department in a large teaching hospital. Emergency nursing students based at a rural university receive that feed, had it convert on the fly to 3D for use within their virtual learning environment. Students may actually ‘work’ a full shift virtually, needing to respond to the challenges of the changing environment as they occur.

As I said, there’s a long way to go (for starters, KinectFusion is about surfaces only), but the progress is rapid and exciting. Over to you: what applications could you see this being good for?

Microsoft Kinect on WoW: evolutionary steps

This post comes from Metaverse Health, our sister site devoted to everything related to health and virtual worlds.

Here’s a great video showing the use of the Kinect to play World of Warcraft using a software framework called FAAST. After the initial demo there’s some useful discussion by the University of Southern California’s Institute for Creative Technologies on its potential applications.

Those applications are something discussed here regularly: rehabilitation and physical activity. Imagine the impact of the technology shown in the video for someone who plays a dozen or more hours of an MMO each week? Let alone someone with a chronic disease or multiple lifestyle risk factors.

Anyway, have a look for yourself:

For my interest in clinical simulation, these developments are of particular interest. Truly effective simulation is likely when health practitioners are physically able to replicate tasks. With my crystal ball in hand, I can’t see that being any more than five years away.

Previous Posts