Researchers developed a novel method to create real-time images and videos of hidden and moving objects using radio signals.
Researchers at the National Institute of Standards and Technology (NIST) and Wavsens LLC have developed a novel method to create real-time images and videos of hidden and moving objects using radio signals. The method can help firefighters find escape routes or victims inside buildings filled with fire and smoke, and can also help track hypersonic objects such as missiles and space debris.
“Our system allows real-time imaging around corners and through walls and tracking of fast-moving objects such as millimeter-sized space debris flying at 10 kilometers per second, more than 20,000 miles per hour, all from standoff distances,” said physicist Fabio da Silva, who led the development of the system while working at NIST.
“Because we use radio signals, they go through almost everything, like concrete, drywall, wood, and glass,” da Silva added. “It’s pretty cool because not only can we look behind walls, but it takes only a few microseconds of data to make an image frame. The sampling happens at the speed of light, as fast as physically possible.”
The technique used here uses a small sample of signal measurements to reconstruct images based on random patterns and correlations. The technique has previously been used in communications coding and network management, machine learning and some advanced forms of imaging.
Da Silva explains the imaging process, “To image a building, the actual volume of interest is much smaller than the volume of the building itself because it’s mostly empty space with sparse stuff in it. To locate a person, you would divide the building into a matrix of cubes. Ordinarily, you would transmit radio signals to each cube individually and analyze the reflections, which is very time consuming. By contrast, the NIST method probes all cubes at the same time and uses the return echo from, say, 10 out of 100 cubes to calculate where the person is. All transmissions will return an image, with the signals forming a pattern and the empty cubes dropping out.”
The researchers combined signal processing and modelling techniques to reconstruct images. The transmitting antennas radiate 200 megahertz to 10 gigahertz frequency. The receiver consists of two antennas connected to a signal digitizer. The digitized data gets transferred to a laptop computer and uploaded to the graphics processing unit to reconstruct the images.
The work is described in the journal Nature Communications.
