From: Eugene Leitl (eugene.leitl@lrz.uni-muenchen.de)
Date: Fri Mar 31 2000 - 00:05:52 MST
Mike Steven writes:
> > 20,000,000,000,000, or 20 trillion cameras,
Hey, no problem. We'll just grow them on trees, or something. Or
air-drop them.
> But the vast majority of those cameras would be pointing at a place where
> there aren't any humans, vehicles or robots and so would be worthless. A
Very true. Subways and other form of public transportation where a lot
of people sequentially pass through a tiny crossection, entrances into
buildings/highway systems, street crossings (street lights).
> fairly universal surveillance system (obviously I agree no system will be
> perfect) could get by with a tiny fraction of that number of cameras,
> particularly if the cameras are mobile. As for cost, already today you can
> get an (admittedly rubbish quality) web-cam for less than £50, and at that
A mass-produced solid-state camera including optics should cost no
more than $1, or less. At current state of the art, it would be
expensive to equip each of them with image recognition (especially as
the current state of art is both puny and advances so fast), so you'd
set up distributed DSP surveillance centers.
As to limited resolution: unsharp masks. Gait, gesture and movement
pattern evaluation do not require a lot of pixels to provide good
results. Plus, zoom optics and automatic tracking. You don't have to
track the whole crowd simultaneously after you've found your suspect,
have you?
With sufficient snoopware density, you can track a suspect across wide
areas. And of course things grow very interesting if one starts to use
data warehouses to crosscorellate data gathered from multiple
sources. Very interesting indeed.
Then there will be speech recognition, chips which do on-board
neuronal OCR, body odor detectors, tags including high-resolution
positioning detectors, the whole gamut of it.
> price it wouldn't be exhorbitantly expensive to have every room in your
> house under surveillance. (You may be relieved to hear that I don't have any
> plans to start webcasting Meonthetoilet.com any time soon :)
Bummer. And we were hoping ;)
> If that level of surveillance is possible today, I don't think it's a huge
> stretch to postulate a future system which could cover not just your house,
> but also shops, businesses, public areas etc and at significantly higher
> quality than the best available today. Sure, if you decide to go hiking in
> the middle of nowhere there won't be any cameras, apart from those you bring
> with you (or ones which follow you if we soup up the tech a bit).
LEO stuff can sure track a human moving across a landscape. And
helicopter based optics (passive infrared and optical) makes person
tracking 1) unavoidable 2) undetectable.
With nanotechnology one can make surveillance really 100%.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:27:46 MST