Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed video magnification algorithms that make virtually imperceptible changes such as movements and color variations visible to the naked eye. The team initially developed the technology to measure heartbeats from a video stream by detecting the subtle changes in skin color caused by each pulse. They have subsequently used it for numerous other applications, including studying a video of a vibrating object to reconstruct ambient sound or estimate the object’s material properties.
The CSAIL team used MATLAB® to develop, refine, and deploy their video magnification algorithms.
“Like many research teams and universities, we value the ability to rapidly test ideas,” says Michael Rubinstein, CSAIL research affiliate. “With MATLAB I can quickly write a prototype algorithm and see if it works. I can then share it with students and collaborators to further build on and refine the prototype.”