When it comes to security cameras, they are only of any good if there is someone watching all the goings on, and making the appropriate response should a crime be caught in progress. Alternatively, just like in any good action movie, the good guys who want to break into a high tech facility have a dozing security camera person manning the post, going to prove once again that a chain is only as strong as its weakest link. MIT’s researchers figured out that there had to be a better way to get things done, and they developed a smart computer system that functions like a human detective, capable of identifying possible intruders automatically – so it does not matter that James has gone out for a smoke, to relieve himself or just to chow down on a donut at the pantry.
This MIT system is capable of calculating just how to best scan footage from security cameras, whether it ought to rely on skin detection algorithms first to ID a person, or to kick off with background detection to spot unusual objects. In short, it is a dynamic, thinking system that will make a call (and hopefully the right one) when the time comes, running tests first to learn just how fast each method takes in a a particular surveillance scenario.
According to Christopher Amato, a computer science postdoctoral researcher at MIT’s Computer Science and Artificial Intelligence Laboratory, “Sometimes it’s important to come up with an alarm immediately, even if you are not yet positive exactly what it is happening. If something bad is going on, you want to know about it as soon as possible. You can’t have a person staring at every single screen, and even if you did the person might not know exactly what to look for. For example, a person is not going to be very good at searching through pages and pages of faces to try to match [an intruder] with a known criminal or terrorist.”