In the previous post I mentioned how critical is to have the clearest images as possible to analyze and for this reason is important to know all the tricks to set up the environment. The most important trick I reckon is knowing when and how to use and InfraRed (IR) camera. In facts, these kind of camera react very differently than our eyes because they ignore some light spectrums. This allows you to get very readable images from contexts you would never expect. Not only IR cameras can be used for night-vision but in Computer-Vision context they can be used to ignore completely the visuals of projectors or normal displays so that you can clearly recognize people interacting with screens.
InfraRed cameras are usually used for security proposes and are usually combined with IR lamps. They also have in general as output a composite video that has then to be converted in digital to be analyzed. For this reason the most common solution is hacking a webcam and replacing its IR filter with something which can block visible light (nowadays there is an incredibly variety of webcams, from the very cheap to the ones with high resolution or with wide angles).
The process is very simple and you can follow it in this tutorial. I was a bit concerned on how to find a developed film with black parts but the solution was easy. If you don’t have it yet somewhere, go to a photo shop, buy a new color roll film for camera, open it so that it gets totally exposed (all the pictures would be white) and then make them develop it (being negative the film returned will be all black), it’s better you let the shop assistant know your aim so that you avoid them to generate the pictures and return you only the film.
The following picture shows a capture from my camera after being hacked. As you can see the images shown in the screen, probably my webmail, is totally ignored and with a simple background subtraction my finger can be identified clearly.
