Boston Police Chief: Facial Recognition Tech Didn't Help Find Bombing Suspects

A reader writes "ArsTechnica reports: 'While the whole country is relieved that this past week's Boston Marathon bombing ordeal and subsequent lockdown of the city is finally over, Boston Police Commissioner Edward Davis told the Washington Post that the department's facial recognition system "did not identify" the two bombing suspects. "The technology came up empty even though both Tsarnaevs' images exist in official databases: Dzhokhar had a Massachusetts driver's license; the brothers had legally immigrated; and Tamerlan had been the subject of some FBI investigation," the Post reported on Saturday. Facial recognition systems can have limited utility when a grainy, low-resolution image captured at a distance from a cellphone camera or surveillance video is compared with a known, high-quality image. Meanwhile, the FBI is expected to release a large-scale facial recognition apparatus "next year for members of the Western Identification Network, a consortium of police agencies in California and eight other Western states," according to the San Jose Mercury News. Still, video surveillance did prove extremely useful in pinpointing the suspects.'"

Wasn't It As Much Individual Photog & ID?

[bc90021]

If I'm not mistaken, the CCTV footage was not as useful... what did help was the one man who took a picture of the bomber (unbeknownst to him at the time), and more importantly, the unfortunate man whose legs were blown off at the knees who valiantly gave an ID from his hospital bed.

Re:CCTV

[paiute]

The MIT policeman was apparently shot in the head while sitting in his car, not apprehending the suspects.

Re:Enhance it and zoom in

[kannibal_klown]

What? You mean you can't "Zoom and enhance" a 640x480 video to the point that you can see the fingerprint left on a window 25' away

Unfortunately, I know people that actually think that stuff is legit. Which of course leads to "fun" arguments / questions about "Why can't you do THIS, I see them do it on TV all the time."

Re:Enhance it and zoom in

[plover]

Actually, even low resolution cameras can be really useful, under certain conditions. If the suspect stands still for a few frames, the images can sometimes be enhanced due to motion differences between the frames. The process is like anti-aliasing in reverse.

In the video clips i saw on the news the suspects were walking, and the differences between frames looked too great to get the kind of data needed to interpolate.

If you're interested in seeing this done in a non-fakey-CSI application, Thierry Legault is an astrophotographer who uses frame interpolation to produce amazingly clear shots of objects like the ISS. See his site here to learn more: http://legault.perso.sfr.fr/ [perso.sfr.fr]

Re:Wasn't It As Much Individual Photog & ID?

[interkin3tic]

Just to note, Bauman (the guy we're talking about, lost legs, IDed bomber) DOES have health insurance [thinkprogress.org]

Not only does Bauman have employer-sponsored health coverage through Costco — the company “is also matching donations made by colleagues at the chain’s Nashua location,” according to a more recent Globe article from Friday. Bauman is being forced to raise funds despite this assistance due to the extraordinarily high costs associated with the amount of current and ongoing care that he requires.

Personally, I think this is a perfect example of why having health insurance run by for-profit organizations is a terrible idea and why the taxpayers paying for health insurance would be better. Anyway, the victims are being taken care of better than most citizens will be, as of friday, three had sites where people could donate to their health costs, and they were all above $400k. In at least Bauman's case, his employer is matching, so that's more like $800k.

Re:Enhance it and zoom in

[Psyborgue]

There are actually some decent enlargement algorithms [weizmann.ac.il] out there that can do better than you would probably think possible. Take a look at the vision chart image. GIGO is an unbeatable law, but there's a lot more you can do with "bad" input than most geeks think. Dude below talking about license plates has the right idea. If the missing information is present elsewhere in the image, or if you have prior knowledge, you can use that to reconstruct portions of the image you want. The more repeating patterns in the image, the better.