What happens if you apply some more advanced techniques that can be found in the autonomous cars to other disciplines? Alex Teichman, of Stanford University, designed a system for cars of the most peculiar in 2012: is able to detect what is happening around it Although this may change. For example, if there is a cyclist in the vicinity, any car with image analysis would see it. But and if the rider removes his hands from the handlebar and changes of posture? Most algorithms possibly not detected during it as cyclist.
This changes with your system: car you know that before he had a cyclist there because she saw him previously and was it switchover, so now you know that it continues to be a cyclist. And if this applies to security systems? That is precisely what he has done in his new prototype. What happens if there is a branch opposite the front window and moves? The alarm would not jump because the system already knows that the branch was there, and that simply has changed or is changing shape or position. And if a thief scale the facade? Then it does detect it.
Thief crawl or dog?
The same applies if you have a dog at home, as you can see in the video that we leave on these lines. The dog was already there and it is not a threat, so it is not cause for alarm Although it moves or jump around the room. If someone comes crawling and trying to imitate the behavior of the dog, on the other hand, the system that detects it and marks it as threat.
The idea of Teichman is improve the traditional image analysis system and replace it with a computer that reconozcalos follow as they move, learn more about what surrounds him in the process. By now it is of a simple prototype, It is not for sale and you in fact do not have name, but its creator has in mind start to hire engineers for your company over the next few weeks in order to commercialize his idea in the market.