Vision System on Chip Thinks for Itself

BALTIMORE, Md., March 22 -- A Johns Hopkins University electrical engineer has developed a vision system on a chip that someday may allow a robotic surgical tool to locate and operate on a clogged artery. The chip performs analog and digital processing, extracts relevant information, makes decisions, and communicates them to the robot.
Putting sensing and processing functions in the same place is called computational sensing, said Ralph Etienne-Cummings, an assistant professor at Johns Hopkins. By processing and reacting to light as soon as it hits the system, the sensors mimic the early type of processing that takes place in an animal's eye. In nature, explained Etienne-Cummings, animals see and react very quickly to the size of objects before them; they approach those small enough to be potential food, and avoid those large enough to pose a threat.