Two years ago, researchers at MIT used Wi-Fi signals to see through walls and track a person's movements. The same team has now come up with a new technology that's not only able to discern a person's silhouette through walls, but can also make out different individuals.
Called RF Capture, the device works by sending and receiving wireless signals that travel through walls. When the wireless radiation encounters a human body, it's reflected off them and the received signal is captured and analyzed. The radiation used, according to the team, is quite minimal, roughly 1/10,000 of the radiation given off by a standard cellphone.
Since the same signal bounces off all the different parts of a person's body, the team had to tackle the tricky issue of accurately processing the received signal in a way as to differentiate between various body parts. This was made harder by the fact that the researchers could only get a little usable data from these reflected signals.
"However, we can extract meaningful signals through a series of algorithms we developed that minimize the random noise produced by the reflections," says Dina Katabi, Director of the Wireless@MIT center.
The device first makes 3D scans of the entire space to identify objects in the environment, including the people in it. When people move about in the environment, the device monitors the signals reflected off their body. It then stitches these reflections across time to reconstruct a person's silhouette. To distinguish between different people, the team trained the device to use factors like body shape and height to create specific "silhouette fingerprints" for every person.
Currently the device is reported capable of being able to trace a person's hand as he writes in the air, through walls and closed doors. It can differentiate between 15 different individuals with a 90 percent accuracy rate.
Potential applications for the technology include gaming (envision being able to interact with a single game from different rooms or initiate actions based on hand movements), emergency response and even motion capture in filmmaking.
"RF Capture would enable motion capture without body sensors and could track actors’ movements even if they are behind furniture or walls," says PhD student Fadel Adib, the study's lead author.
Smart homes or caring for seniors are other areas where the device's motion-capturing abilities could potentially come in handy.
"We’re working to turn this technology into an in-home device that can call 911 if it detects that a family member has fallen unconscious," says Katabi. "You could also imagine it being used to operate your lights and TVs, or to adjust your heating by monitoring where you are in the house."
Currently the team is in the process of commercializing the technology through a product called Emerald, designed to detect and prevent falls among the elderly. Emerald was presented to President Obama as part of the White House’s first annual Demo Day a few months ago.
The team's paper (PDF) on RF Capture was accepted to the SIGGRAPH Asia conference and exhibition currently taking place at Kobe, Japan.
Source: MIT
