Motion amplification

A press release from MIT describes computer vision/image analysis work on detecting subtle frame-to-frame variations. Along with an impressive video, some interesting application areas are proposed.

Chief among these is the idea of non-contact heart rate monitoring, such as in hospital neo-natal units, but my thoughts are not quite so noble. I wonder if this technology – described as “motion amplification” – has process control applications.

Could we for instance, watch a fluid passing through a solid tube, or how about monitoring a paper web or cotton thread? In all of these, small changes could be a precursor to some larger failure, so early detection could save both time and money.

Or another idea: the MIT researchers say, “if the range of frequencies is wide enough, the system can amplify changes that occur only once. So, for instance, it could be used to compare different images of the same scene, allowing the user to easily pick out changes that might otherwise go unnoticed.”

In which case, could this technique be used to inspect woven materials? If it were possible to detect subtle changes in the thread or the woven material a lot of waste might be avoided.

In short, interesting work that I think has application in the world of machine vision.