home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Illusion - Is Seeing Really Believing?
/
Illusion - Is Seeing Really Believing (1998)(Marshall Media)[Mac-PC].iso
/
mac
/
ILLUSION
/
SROCK_TX.CXT
/
00111_Text_rel04t.txt
< prev
next >
Wrap
Text File
|
1996-12-31
|
2KB
|
60 lines
How does the perceptual
system distinguish luminance
edges from illumination edges?
We do not yet have a complete
answer to this question, but the
perceptual system may make
use of several kinds of
information available to it. The
best known is the gradual
transition from dark to light
that usually accompanies
illumination edges, namely the
penumbra around cast shadows.
A penumbra occurs because the
sun, an incandescent bulb, or
another source of light is an
extended region, not a single
point. In the previous
photograph, the shadows
running across the palm leaf
against the direction of the
folds are an exmple. The
graduated change in luminance
at the edges of these shadows
are penumbras. They enable us
to identify the shadows as such,
thereby enabling us to perceive
the actual color of the
underlying surfaces. Another
probable indicator is a
difference in the orientation of
neighboring planes or surfaces,
as in the palmΓÇÖs folds. It is
generally the case that the
illumination will be unequal
across folds and corners. Given
the tendency to interpret a
luminance difference at a
corner as based on
illumination, we can now
appreciate why constancy
occurs in such cases but not
when we perceptually "flatten
out" that corner. A third
possible indicator is the
magnitude of the luminance
ratio at an edge. A lightness
edge can be no greater than
about 30 to 1 because that is
what we get from a white with a
reflectance value of 90% and a
black with a reflectance value
of 3%. But an illumination edge
often yields a very high ratio
because the only constraint on
its magnitude is light indirectly
reflected back on the
unilluminated surface by all
other surfaces.