Colour appearance of perceived illuminants and surfaces
Abstract
The colour appearance of a patch of light is only partly determined by the
light itself. A patch's perceived colour can also be influenced by cognitive
processes. Cognitive effects on colour appearance were examined by
modifying an observer's perceptual representation of surface-ambiguous and
textured patches of light. Observers were instructed to view a 1° square test
patch as either an illuminant source or as an object reflecting light from its
surface. The test patch was presented on a high-resolution Nanao 9080i
colour monitor driven by a 32 bit microprocessor (T1 34020 GSP) specialized
for graphics operations. The luminance of the phosphors was
photometrically calibrated and linearized. The five test conditions were a(n):
(1) isolated homogeneous test field (ambiguous test); (2) isolated test
containing a random-dot speckled pattern (textured test surface); (3)
homogeneous test with homogeneous surround (ambiguous test and
surround); (4) test containing a similar texture surrounded by a
homogeneous background (textured test surface with ambiguous surround);
(5) textured test with a textured surround (textured test and surround surface).
Changes in colour appearance were assessed using a red/green hue
cancellation technique based on opponency mechanisms in the human
visual pathway. The test was an admixture of "red" (Xd = 625 nm) and
"green" (kd = 535 nm) light. The chromaticity of the backgroimds were "red"
(Xd = 625 nm) and "green" (Xd = 535 nm). The test patch was varied across
five luminances (0.92 to 19.9 cd/m^). Observers monocularly viewed the CRT
with their left eyes, and were required to adjust the radiance of the "green"
component so that the test appeared neither reddish nor greenish. Results
showed that all but one observer viewed the CRT images independently of
instruction set. The observer showed a greenness (redness) shift in test colour
appearance for surface (aperture) mode perception when viewing a
homogeneous test with a 625 nm surround. The remaining observers
showed either no shifts or shifts in colour appearance that were in the same
chromatic direction as the surround. The data can best be described by neural
assimilation processes.
Collections
- Retrospective theses [1604]