MIT Researchers Pick On Variations, Reveal The Invisible In Video

Scientists at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) are ready with their entry for this summer's Siggraph, the premier computer-graphics conference, with a software that can read and amplify the variations between the frames in a video to reveal objects that go unnoticed to the naked eye. For example, the software can actually discern the "pulse" of a person, by heightening certain frequencies which expose tiny motions.

#-Link-Snipped-#

The software allows the user to set the desired range of frequency and the degree of amplification rendering original and altered version of the video simultaneously in real time. The software is more suited to outline occurrences that have a tendency to recur but given a wide range of frequencies, the system is capable of amplifying instances that occur just once. This quality can be used to compare different images of same scene to single out the changes.

Graduate student at MIT, Michael Rubinstein, who designed this software along with his colleagues, foresees that the system could be useful as an application for "contactless monitoring" of hospital patients' vital signs. Other applications include baby monitors, laparoscopic imaging of internal organs and long-range-surveillance systems.



Source: #-Link-Snipped-# Image Credit: #-Link-Snipped-#