From: Andrew Clough (aclough@mit.edu)
Date: Wed Feb 12 2003 - 10:53:13 MST
At 08:52 AM 2/12/2003 +0100, you wrote:
>avatar wrote:
>>With an inksuit covered in digital ink and some MEMS cameras inbuilt into
>>the suit, along with a very chunky computing ability, something like the
>>suit in GHOST IN THE SHELL etc. could be achieved, probably within 5 years
>>or so.
>
>
>The simplest way would probably be if each 1 cm**2 "pixel" also was a
>single pixel camera, that was in turn connected to a pixel on the opposite
>side.
>
>Usually camouflage don't have to be perfect to work. Naturally it makes a
>big difference if it is used in an urban or a rural enviroment.
However, if the cloaked person isn't to maintain a constant orientation
with regards to whoever their sneaking past, the cloak will have to decide
*which* pixel on the other side to replicate. Looking at an example:
3
2
4
"A"
1 "C" 5
8
6
5
"B"
C's invisibility cloak should project the input from spot 4 onto 8, but in
order to hide from A you would have to project the input from pixel 6 onto
8. This would be quite a bit of computation; especially when you consider
that 6 and 8 are sending different parts of their respective fields of
view. This is not to say that a true or one point of view invisibility
cloak isn't possible, just that we'll have to wait a little longer. What I
do expect to see real soon is active camouflage, which matches its pattern
to the environment of the person wearing it. Crawl on grass, it selects
the "grass" pattern and matches shades with the surface you are on. Crawl
onto a freshly paved road and it will turn black. Press against a concrete
wall and it will choose "Urban Camo #314" and adjust the shade (It can't be
flat gray or the human shaped shadow could be picked out on the wall,
matching the intensity of the sun to prevent a shadow would probably take
too much power, though maybe it could work indoors).
This archive was generated by hypermail 2.1.5 : Wed Feb 12 2003 - 13:44:07 MST