MIT video analysis experts have developed a new way to amplify subtle shifts in color and motion that are normally invisible to the naked eye.
The result of their work: video that could be used to accurately detect pulse rate based on the human face’s rhythmic flush, monitor babies’ breathing or study the movements of buildings, cranes and mechanical devices.
“You can think about what we’ve made like a microscope, except for video,” says doctoral student Michael Rubinstein, whose team came up with the now patented analytical process they call Eulerian video magnification. “It’s a tool to amplify small spatio-temporal variations you can’t normally see.”
See the video after the jump.