From: Robert J. Bradbury (bradbury@www.aeiveos.com)
Date: Sat Sep 11 1999 - 15:48:24 MDT
On Sat, 11 Sep 1999, Ian Goddard wrote:
> So I
> watched the FLIR, which rotates around the
> Center dozens of times, and there are *never*
> any sunlight reflections in the water from
> any angle. This is an ideal reflection-test
> that basically proves that the FLIR did not
> pick up sunlight reflections that a regular
> camera did. This makes sense if you think
> about it. The beam of sunlight coming off a
> body sitting in the sun is not itself going
> to be a lot (or any) hotter than the body
> itself, and the FLIR only sees hot/cool.
>
> I believe that that observation does fatal
> damage to the sunlight-reflection theory
> regarding the obvious gun-muzzle flashes.
>
I think you need to study some physics about
light, heat and reflections.
The infrared enegy coming off of an object sitting
in the sun can be *alot* greater than the infrared
energy it receives. The object absorbs the visible
light energy and re-emits it in the infrared.
Why do you think black tar or sand gets *hot*?
The amount of IR emitted by the object is going
to be proportional to its absorbance of other
frequencies.
This gets confused by reflections which depend
on the angle of the light source to the material
and the reflectivity (complement of absorbance).
The only case you would get a reflection from the
water is if the angle of the sun to the water
to the camera was precisely the angle required
for a reflected image. If the sun is directly
overhead or at a low angle, it is unlikely that
you would see any reflections.
The only case a "beam of sunlight" comes off
of an object sitting in the sun is when that
object is a mirror or happens to be ~white.
All other color objects reflect on the color
they appear to be and absorb all other colors.
See:
http://www.scitex.com/solutions/colormng/color1wp.htm
I would leave the determination of reflections vs.
gun muzzle flashes to people who are trained to
interpret FLIR images.
Robert
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:05:07 MST