Radars are not known for having a keen sense of detail. They shoot out radio signals and measure the reflections, rendering an object's location as a series of digital coordinates or a pulsating dot on a radar display. Cameras or laser scanners are better equipped to determine an object's shape or size.
But in recent years, new advances have improved radars to the point where they can generate rough images of the physical world. Several companies are making radars that help drones and autonomous cars avoid obstacles using specialized materials and software.
Now researchers in Scotland have further enhanced what radars can do – using radar to instantly recognize objects like metals and body parts and distinguish between them, according to a paper presented last month at the Symposium on User Interface Software and Technology.
To build the classification system, the researchers tapped into a miniature radar chip that Google unveiled at its I/O developer’s conference in 2015. The chip, called Soli, had been designed to track slight finger movements, so that people can use hand gestures to control computers and smartphones.
But the researchers at St. Andrews University found that the radar chip can recognize slight variations in the surface and composition of objects placed against it. The radar signals reflect off objects and materials in unique ways, creating something like a digital signature.
The system, also known as Radar Categorization for Input & Interaction or RadarCat, trains itself with machine learning algorithms to read those signatures and assign them to an object. It has been shown to instantly identify things like sponges and smartphones, differentiate between copper and steel, and tell if a glass of water is empty.
