Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-l] An Unsolved Color Problem



On 07/17/2007 08:08 AM, Polvani, Donald G. wrote:
Here is the Astronomy Picture of the Day (APOD) for today July 17,2007.

http://antwrp.gsfc.nasa.gov/apod/astropix.html

I haven't "solved" the problem of how the two squares can appear to have
different colors and yet be the same color,


The human eye is not a spectrophotometer!

Yeah, I know every grade-school science book implies that the eye
just responds to the colors of the incoming light, pixel by pixel
... but it just ain't true. Not even close.

There's no reason why it should be true. Given a few million years
of primate evolution, you would expect the eye to evolve to so that
it can see the difference between ripe fruit and unripe fruit, and
see the difference between the tiger and the tall grass. Doing a
good job of this requires _color constancy_ ... where color refers
to the inherent pigmentation of /objects/ in the field of view and
requires _discounting the illuminant_ because the illuminant is
certainly not inherent to the objects, and because the illuminant
changes wildly from situation to situation.

http://en.wikipedia.org/wiki/Color_constancy