h a l f b a k e r yCall Ambulance,
Rebuild Kitchen.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Imagine a security camera, looking at a large street scene. To get high-res pictures of people in this scene, the camera would have to be very
high-spec, expensive, with a
high-quality lens and would produce large volumes of data.
This idea is to replace the single high-res camera with two
cheap, low-res cameras. One camera is static and views the whole scene and the other has a longer lens and thus a small field of view sufficient to show a person in reasonable detail. This second camera moves automatically to point at moving objects in the scene imaged by the first camera.
The first camera produces a stream of
low-res video. The video from the second camera is inserted into this video stream at the right coordinates. It'll look a bit MPEG-ey and blocky but the result will be a
low-resolution video with moving people shown in high-resolution. Total camera costs should be lower and file sizes will be much smaller.
Two Cups of Coffee
Two_20Cups_20Of_20Coffee
[normzone, Feb 20 2012]
What are bollards?
http://www.google.c...t=i&cr=&safe=images
[normzone, Feb 20 2012]
bollards
http://www.flickr.c.../72157622499908406/
[mouseposture, Feb 20 2012]
[link]
|
| |
somebody's been watching "Person of Interest". |
|
| |
Probably, but not me - I've never heard of it |
|
| |
you've pretty well described the opening sequence: low resolution area coverage dotted with high-res faces/lic.plates. |
|
| |
Anyways, what do you do in the case of multiple subjects ? |
|
| |
Would it not be cheaper yet to have a single, fixed
high-resolution camera, but transmit only the parts
of the image that are of interest (eg, moving)? That
would achieve the necessary saving in file-size or
transmitted data, and would be simpler and (I
suspect) a lot cheaper. |
|
| |
(I suspect that digital TV is transmitted in a roughly
similar way, with only the changing parts of the
image being sent, no?) |
|
| |
[Max] I think that's right - digital video is encoded using logic similar to that. Then it comes down to whether two cheap, simple cameras are cheaper than one high-quality, high-definition camera.
[FT] Alternate between the two objects - so each will be in high-definition for some of the time. |
|
| |
Ninjas would learn a new technique: instead of
blinding speed, they'd move with glacial slowness,
and the security cameras would show only a blur. |
|
| |
If the only outcome of this innovation is to fill our streets with slow-moving ninjas then it will have been worthwhile. |
|
| |
Maybe that's what the Falun Gong is really about. Perhaps the Chinese Government has already invoked this policy... |
|
| |
If you watch CSI with any regularity, you will realize
that even cheap camera images can be 'enhanced' by
computer power to resolve faces from miles away,
license plates in the next city and have the ability
to render images even in near darkness. |
|
| |
And DNA identifacation can be completed within hours of
the crime being commited, thus leading to nice, tidy,
same-day arrests that tie things up before the credits roll. |
|
| |
" If you watch CSI with any regularity, you will realize " |
|
| |