I know, it sounds like something out of a Sci-Fi movie where people
are arrested for crimes before they commit them. Oh wait, that movie
was already made. It was not a good idea onscreen, so researchers decided to find out if it was a good idea in real life. Well, not exactly, but they are getting one step closer to the reality of a time when a computer decides if you are suspicious.
Engineers are developing a computerized surveillance system
that, when completed, will attempt to recognize whether a person on the
street is acting suspiciously or appears to be lost. The first three phases of the project are already done:
they have one software algorithm that creates a wide-angle video
panorama of a street scene, another that maps the panorama onto a
high-resolution aerial image of the scene, and a method for actively
tracking a selected target.
So does this mean that the machines will have to know who you are?
According to the creators of the device, no, this will not require any large database of citizens or knowing who the person is to figure out what they are doing. "In our research, we care what you
do, not who you are. We aim to analyze and model the behavior patterns
of people and vehicles moving through the scene, rather than attempting
to determine the identity of people. We are trying to automatically
learn what typical activity patterns exist in the monitored area, and
then have the system look for atypical patterns that may signal a
person of interest -- perhaps someone engaging in nefarious behavior or
a person in need of help."
This system is more accurate than the traditional camera methods for
several reasons. When surveillance operators look through a traditional
camera they get only a tiny image -- what some refer to as a "soda
straw" view of the world. If you have ever seen the opening to a James Bond movie then you know what that field of vision looks like. As they move the camera around, they can quickly lose a sense of where they are looking within a larger context.
The Ohio State software takes a series of snapshots from every
direction within a camera's field of view, and combines them into a
seamless panorama.This is similar to the kind of photo stiching software that you find on a few online photo sharing sites. The places where the images overlap are eliminated and only one seemless image of the whole area remains.
How does the camera create an accurate panorama? After all a passing bus could easily obscure a large area and make accuracy in real time a challange. Well, the researchers though ouf that. This software
maps locations onto an aerial map of the
scene, such as a detailed Google map.
A computer can use this information to calculate where the viewspaces
of all the security cameras in an area overlap. These images can them be used to make a panorama that is accurate.
If this all seems a little bit 'Big Brother' to you just wait. The computer is going to try and judge your intentions.That right, computers, those things without feeling, will be trying to judge your feelings by your body langauge. It will then decide if you are shady.
To first determine what constitutes normal behavior, they plan to
follow the paths of many people who walk through a particular scene
over a long period of time. A line tracing each person's trajectory
will be saved to a database. "You can imagine that over a few months,
you're going to start to pick up where people tend to go at certain
times of day -- trends," one of the researchers said. People who stop
in an unusual spot or leave behind an object like a package or book bag
might be considered suspicious by law enforcement.
Mimes watch out. You may be pegegd as a terrorist by the computer
when you do your sneaking immitation or your wind walk. While this
system has potential it lacks the one thing that humans have, context. Clearly, it will still need the human touch.