A system that allows for high-accuracy 3D motion tracking could change the way gamers are able to interact with their real-world environments by detecting radio signals reflected off their bodies -- even through walls.
The WiTrack system has been developed by a team at MIT's Computer Science and Artificial Intelligence Lab (CSAIL) can pinpoint location within 10-20 centimetres and track both 2D and 3D movement using specialised radio signals. Four antennae are used -- one for transmitting, three for receiving -- to track the radio waves. A gemoetric model of the user's location is built based on the distance between the antennas and the user.
The WiTrack is a refined version of a system called the WiVi, which was built earlier this year by two of the team's members and is also similar to a device called the WiSee, created at the University of Washington. Instead of radio signals, the WiVi and WiSee both relied on Wi-Fi, but because the radio signals are around 100 times smaller than Wi-Fi signals, they allow for faster and better location accuracy. The WiTrack is also able to eliminate reflections off walls and furniture, and concentrate solely on human motion.
The primary use for the WiTrack as envisaged by the team is for the WiTrack to be used to allow gamers to explore virtual reality environments with real movements, such as running down corridors and hiding from enemies. This has been achieved already to a certain extent with omni-directional treadmills such as the Virtualizer and the Omni, but would it work with real-world environments, where their will almost always be a physical disparity between the on- and off-screen spaces?
"We believe there are multiple options for matching the players to the spaces," Fadel Adib, a graduate student at CSAIL, tells Wired.co.uk. "One of them is the goggles, glasses, or virtual reality headsets like Google Glass, Augmate, Space Glass, or Oculus Rift. Some of these headsets are transparent, so they can overlay the game on top of the background, and enable the players to avoid real-life obstacles, or even better, try to image these obstacles using their sensors and incorporate them into the games in real-time."
It could also be used in a bigger space, such as Microsoft's Illumiroom, he suggests, or as a temporary and rudimentary option, furniture could be rearranged in a room to match the in-game environment.
"We believe that game developers are very creative, and so exposing the through-wall 3D tracking functionality to them would enable a generation of games we never thought were possible and are beyond what we could imagine today," he says.
Other possible applications of the WiTrack include monitoring falls of elderly people and controlling household appliances. Current fall detection systems requires users to wear sensors or depend on cameras installed in the home, but the team claims that the WiTrack can detect falls with 96.9 percent accuracy just using reflections.
Similarly, because the WiTrack can detect movement of individual limbs, not just whole bodies, it can be used to operate household devices -- by pointing to turn lights on and off, for example -- even if they are in different rooms.
Currently the system can only track one person at a time, but the team are working on improving it so that it can keep tabs on multiple people. As the WiTrack is neither expensive nor difficult to develop, the developers believe it should be easy to bring to market. The CSAIL team are due to present their paper, 3D Tracking via Body Radio Reflections, at the Usenix Symposium on Networked Systems Design and Implementation in April 2014.

0 comments:

Post a Comment

 
Top