NASA catches molester, can't find holes in shuttle

July 16, 2003 1:57 PM

NASA laboratory enhancement of the surveillance tape used to nab suspect in West Virginia Target sexual assault. It's nice to know NASA can use image enhancement technology to catch child molesters, but can't use it to find holes in space shuttles.
posted by dayvin (12 comments total)

A stupid, baiting post.
posted by the fire you left me at 2:12 PM on July 16, 2003

dayvin, you're a moron.

are you happy now?
posted by shadow45 at 2:23 PM on July 16, 2003

I have to agree...what a stupid way to phrase this post.
posted by dejah420 at 2:28 PM on July 16, 2003

cuz holes in the shuttle are rare, and besides, as the second link tells it, the photography is insufficient. on the other hand, everybody knows that collective armies of child molestors are slinking about major malls and elementary schools, breathing heavily and salivating over the opportunity to violently deflower our precious children. fortunately, virtually every populace public place is now under constant video surveillance. we're good.
posted by quonsar at 2:47 PM on July 16, 2003

It's obviously another case of someone being railroaded. Who'd believe that this man could be a child molestor?


posted by Mayor Curley at 3:30 PM on July 16, 2003

mayor curley can detect a child molester just by looking at a person. evidence, due process, trial, conviction - all bureaucratic moneywasting flummery. shhhheeeaaaaah. perhaps the good mayor would inform the rest of us what to look for: is it the short hair? impaired vision? the loose collar? wait - it's the blue eyes, right?
posted by quonsar at 3:50 PM on July 16, 2003

Quonsar,

Can you please elaborate as to whether you're peeved because you're taking my comment at face value, or if you're incensed because I implied that the guy in the photo looks creepy enough to fit a stereotype. I'm not sure how to respond.
posted by Mayor Curley at 7:10 PM on July 16, 2003

Looks like a shop teacher to me.
posted by Witty at 7:15 PM on July 16, 2003

I'm not sure how to respond.

wit and sarcasm are two good choices. :-)
posted by quonsar at 9:11 PM on July 16, 2003

Since nobody's brought up the science being used here, I'll mention it. It's called temporal supperresolution, and the basic idea is to trade the large number of low resolution samples for a single higher resolution image -- based on the idea that each of the low rez images contains a slightly different view of the target.

Methods for doing this vary -- some simply try to filter out sensor noise, others actually achieve subpixel accuracy based on the precise timings of sensor change elicited from a moving object. I'm sure NASA has some particularly wild stuff for this.

A paper on the topic can be found here:

http://www.ece.gatech.edu/research/labs/MCCL/pubs/dwnlds/bahadir_june2002.pdf

Code to implement motionless cleanup can be found here:

http://auricle.dyndns.org/ALE/

I had a link to an intriguing approach using MPEG Motion Vectors to do this kind of work, but I can't find it.

--Dan
posted by effugas at 9:32 PM on July 16, 2003

Low blow, dayvin. Very low.
posted by holycola at 7:51 AM on July 17, 2003

Supperresolution = Really hi-def digital photos of the dinner table?
posted by techgnollogic at 11:26 AM on July 17, 2003