Facial Recognition Software

When Facial Recognition Software was mentioned as a possible way to catch the person or people behind the Boston Marathon bombings, I decided to do some research and write a blog post about it.

Facial Recognition Software basically looks at individual faces in a crowd and compares certain facial features to a database of images already stored.

To do this it measures the different features of a person’s face.  The better software is supposed to use “3D” to make it more accurate.

crowd of people

It’s usually used in security and law enforcement, and as in the Boston Marathon bombings, it’s used to look at random faces in a crowd.

The US government has used it for foreign visitors obtaining a visa.  Their photo is matched up with a database of known criminals and suspected terrorists.

If perfected enough, it could also be used for screening airport passengers, at ATM locations, and by banks to verify a customer’s identity (check cashing, etc.) and prevent fraud.

The TSA (Transportation Security Administration) had a “Registered Traveler Program” that allowed flyers to sign up for an extensive background check system that may have involved facial recognition.

However, it had lots of problems, and it appears it ceased operating in 2009.

Popular Science has a page full of articles on facial recognition software, such as one that talks about software that looks at you and brings up your Facebook profile.

In July 2011, Google bought a facial recognition company called PittPatt (Pittsburgh Pattern Recognition).

The company seems to be top secret.  As of this writing, their webpage only says “Pittsburgh Pattern Recognition has been acquired by Google”, and gives a media contact Email address.

An article at the Computer World site says that Facebook acquired a facial recognition software company (Face.com).

The problem with this technology is that it isn’t perfect.  If a camera can get a straight on front view of a person’s face in good light, the system might work well.

But a lot of things can make it hard on the facial recognition software – faces at an angle, different facial expressions (huge smile, frown, etc.), poor light, long hair, sunglasses, other objects in the way, poor image quality, etc.

There’s also a big privacy concern with facial recognition software.  Some people feel it’s the ultimate expression of “Big Brother”, and are terrified of a technology that may be able to track down your identity in minutes.


Share Button