Colour causes a depth illusion in human visual perception Abstract UNING FUNCTIONS). REPLICATED. EXTENDED KINGDOM 2003. SUCCESSFUL REPLICATION. FOUND THAT SOME THINGS CHANGED WITH ORIENTATION / SF/ PHASE Introduction Scientists are interested in how colour sensitive and luminance sensitive mechanisms interact when a subject is presented with stimuli that embody the particular relationships that exist between colour and luminance in the human vision system. Colours plays a highly crucial role in the vision characteristics of the human sight, because of its immense sensitivity the subject holds great importance among vision scientists throughout the globe (e.g. review by Regan, 2000). The main strategy to gain knowledge in this subject matter is to study the performance, set against a certain criteria, using only iso-luminant (single colour characteristics) and iso-chromatic (multiple colour characteristics) stimuli. The In this study, we aim to study the colour and luminance characteristics of the human vision with an attempt to establish a proper relationship between the colour, depth and luminance in the human vision. Kingdom (2003) proposed that a significant amount of knowledge could be gain by analysing the behaviour of colour and luminance in the human vision perception, and how this phenomenon can emulate the spatio-temporal relationships between the colour and luminance found in human vision. Kingdom et al (2005) attempted the first successful approach to understand this subject and investigate the proper relationship in such phenomenon. They studied that when a chromatic grating is added at a certain level to luminance grating; one of them gains the impression of groovy structure, this process of
18
Embed
Physics lab report sample from assignmentsupport.com essay writing services
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Colour causes a depth illusion in human visual perception
Abstract UNING FUNCTIONS). REPLICATED. EXTENDED KINGDOM 2003.
SUCCESSFUL REPLICATION. FOUND THAT SOME THINGS CHANGED WITH
ORIENTATION / SF/ PHASE
Introduction Scientists are interested in how colour sensitive and luminance sensitive mechanisms interact
when a subject is presented with stimuli that embody the particular relationships that exist
between colour and luminance in the human vision system.
Colours plays a highly crucial role in the vision characteristics of the human sight, because of
its immense sensitivity the subject holds great importance among vision scientists throughout
the globe (e.g. review by Regan, 2000). The main strategy to gain knowledge in this subject
matter is to study the performance, set against a certain criteria, using only iso-luminant
(single colour characteristics) and iso-chromatic (multiple colour characteristics) stimuli. The
In this study, we aim to study the colour and luminance characteristics of the human vision
with an attempt to establish a proper relationship between the colour, depth and luminance in
the human vision. Kingdom (2003) proposed that a significant amount of knowledge could be
gain by analysing the behaviour of colour and luminance in the human vision perception, and
how this phenomenon can emulate the spatio-temporal relationships between the colour and
luminance found in human vision. Kingdom et al (2005) attempted the first successful
approach to understand this subject and investigate the proper relationship in such
phenomenon. They studied that when a chromatic grating is added at a certain level to
luminance grating; one of them gains the impression of groovy structure, this process of
transference from colour to shape is called the Depth Enhancement. On the contrary, if
second grating of chromatic grating is further added to this process at a different level, the
impression of depth is either reduces or completely eliminated, and this process of
elimination or reduction is called Depth Suppression. These kinds of phenomena are
generally experienced in achromatic kinds of studies that are highly influenced by different
colour contrasts (Lehky & Sejnowski, 1988; Ramchandran, 1988; Attick et al., 1996; Sun &
Perona, 1997).
The depth enhancing processes, formation of a shape from shade due to the grating between
chromatic and luminance patterns proposed that natural human visual system has certain
inbuilt capabilities;
1. The main cause of variation in chromatic and luminance behaviour that are spatially
aligned against each other is due to the variation in surface reflectance.
2. The main cause of pure or impure variations in the luminance behaviour is due to the
non-uniform illumination, such as shading and shadows.
These physical relationships between the chromatic and luminance grating holds
dominant importance in the field of human vision and the scientist are contented upon
the agreement that such relationships give rise to acknowledge these in-built system in
the human vision (Rubin & Richards, 1982; Cavanagh, 1991; Mullen & Kingdom,
1991; Olmos & Kingdom, 2004) along with the colour shading affect which is quite
evident to appreciate that these are embedded into the human visionary configuration.
There are several other factors that may signal the perceptions of surface shapes. There is
interest in whether colour contrast on the perception of shapes influences the perceptions of
shading, such as texture; and whether the colour contrast influence the contribution of
shading to surface curvature when it is present alongside other cues. It is possible that the
influence of colour contrast on shape-from-shading is reduced, or even eliminated, when
surface information other than colour is present, because in such circumstances the surface
versus illumination interpretative role of colour contrast becomes redundant. The aim of this
study, as we have already mentioned, is to establish a relationship between the colour,
luminance and depth of human vision with a focus to investigate the influences of the colour
contrast on perceived shapes in pattern that produces shape from shading with shape from
texture. Mamassian and Landy (2001) also noticed that the orientation defined textures have
been shown to combine synergistically with shading to create strong impressions of depth.
Numerous questions have so far been aroused concerning the chromatic properties of the
colour shading. It is often bring into consideration to investigate that the combination of two
phenomena, depth enhancement or depth suppression, in colour directions is more important.
There is high possibility the colour shading effect is weaker when the directions of depth
enhancement and depth suppression phenomena is same, as in such phenomenon the human
vision system might bound together both colouring patterns into a single object, releasing the
luminance variations from being designated as changes in reflectance, and designating them
instead as shading, even though they are spatially aligned with one of colour patterns.
In this study, we have attempted to answer the questions regarding mixed colour and
luminance plaids with an aim to manipulate the direction and colour texturing of both the
depth enhancement and depth suppression. The results of this study furnishes further about
the information and understanding of the chromatic properties in terms of colour shading
textures and formation of shape from shading by the natural human vision system, and
therefore tries to acknowledge the assumptions related to the relationships between the
colour, luminance and depth of the vision system. In order to grab proper and precise
knowledge about the relationship between the colour, luminance and depth in the perception
of vision, we have utilize an adjustable stimulus that have original and real corrugations and
bumps in its structure, defined stereoscopically.
The findings of this study can be summarised by suggesting that the impression of depth is
presented when variations in colour were appeared at different orientation to plaid gratings
and at the same orientation but out of phase, therefore, the colour variations at different
orientation and out of phase will yield the depth enhancement. Additionally, the addition of
colour variations of the same orientation and in phase will suppress the grating that will yield
the depth suppression.
Method
Participants
The participants who took part in the chromatic-achromatic experiment were 8 psychology
students and 1 professor in the University of York: A, E, S, J, R, JS, JR, H and AW. For the
other 5 experiments, there were 8 participants in total except subject A. Details such as age,
gender and handedness were not necessarily collected for this experiment. Personal
identifying information were used anonymous.
Materials
The materials used in this experiment were stimuli viewed though a CRT screen and a
keyboard. First of all, the screen was a NEC Multisync 200 screen and the diagonal size of
screen was 20 inch. The refresh rates of how quickly the screen updates were 100Hz. Figure
1 shows the viewing distance between eyes and the screen. The viewing distance was the
total distance (1+2+3+4+5) about 700mm, but not the straight distance between eyes and
screen. The field of view (w x h) was 29.50 x 22.34 deg. The structure of depth (disparity)
was the distance between fovea and the place of images. Each eye was separated to see the
stimuli, it was the way that how stereo images achieved. The reason for that was to make
each eye construct different figures to achieve the disparity not to change the overall disparity
of objects.
5
4
3
2
1
Figure 1. Lab settings in this experiment
Secondly, participants’ adjustments of the amplitude of the depth corrugations in the stereo-
images were made by pressing the up and down arrow keys on a standard keyboard. The
mean starting value of adjustments were -6.00 and the range of that was between -5.00 and
+5.00. Subjects’ responses were accompanied with auditory indicators.
Thirdly, stimuli were displayed on a grey background. The perform calibration of all
phosphors for chromatic data showed the location in colour space for 3 guns. It was
Screen
Eyes
Mirror
calibrated by using the Spyder 3 colorimeter. The R (red), G (green) and B (blue) gun outputs
were gamma-corrected after calibration. The CIE coordinates of the monitors’ phosphors
were R: x=0.640, y=0.330; G: x=0.300, y=0.600; B: x=0.150, y=0.060. The stimuli were
constructed from three component gratings: luminance modulated gratings, colour modulated
gratings and drift modulated gratings. All achromatic gratings with contrast of 100% had a
spatial frequency of 2 cpd and an orientation of 90 deg. The stimuli were presented in a
circular, hard-edged window. The achromatic gratings were ‘black and white’, and were
produced by modulating all three RGB phosphors in (1,1,1). The colour gratings were ‘red-
green’, and was designed to dissociate the post-receptor chromatic mechanism that
differences the L (long-wavelength-sensitive) and M (middle-wavelength-sensitive) cones.
The colour space of LMS was (1, -1,0). The drift gratings with contrast of 10% had a spatial
frequency of 1 cpd and were alternated to avoid movement after-effects. Alternation had a
temporal frequency of 1Hz. In this condition, horizontal gratings could not be used because
the shift bars from left to right was not seems to be moved. So, vertical gratings were made to
get larger disparity. All gratings were formed from sinusoidal modulations of cone contrast.
In more details, stimuli were made and divided into 6 conditions in this experiment: 1.
chromatic and non-chromatic, 2. in phase (0°) and out of phase (90°), 3. Orientation (0°, 30°,
60°, 90°), 4. Phase (0°, 30°,60°,90°), 5. Spatial frequency (1, 2, 4, 8 x original), and 6.
Drifting and static.
Design
The IVs (Independent Variables) in this experiment were 6 stimuli conditions. The DVs
(Dependent Variables) was the subjects’ adjustments for each condition. The main design
was a between subjects design.
Procedure
The type of procedure used here was called ‘psychophysics. Participants were asked to
estimate the apparent depth of the corrugations in the stereo-gratings on a CRT screen and
adjusted the amplitude of the depth corrugations in the stereo-gratings until they matched the
apparent depth of the corrugations in the test stimuli by pressing the up and down arrow keys
on a keyboard. There was no time limit. Each testing session took approximately 1 hour and
there were 6 individual conditions. During each session, stimuli were presented in a random
order with several practice trails and test trails. In the experiment 1, 2 and 5, participants were
tested 8 test trails for each condition. In the experiment 3 and 4, participants were tested 5
trails for each condition. Some participants experienced fading of images or other possible
adverse effects such as headache or dry eyes during along time of staring at a computer
screen. So, participants were encouraged to let their eyes roam around the stimuli to avoid the
negative influence on the adjustments. Finally, written consent forms were obtained from all
participants.
Results As illustrated in figure 2, average disparity threshold showed that the chromatic condition
(M= 5.11, SD= 3.32) tends to be higher than the achromatic condition (M=2.93, SD= 2.46).
The mean difference between two conditions was 2.17 and the 95% confidence interval for
the estimated population mean difference is between 0.44 and 3.91. A paired sample test was
carried out to show that the difference between conditions was significant (t= 2.885, df= 8,
p< 0.020, 2-tailed).
Figure 2. Mean disparity threshold for chromatic and achromatic stimuli.
As illustrated in figure 3, average disparity threshold of the in phase stimuli (M= 2.90, SD=
2.14) was lower than the out of phase stimuli (M= 4.91, SD= 3.60). The mean difference
between two conditions was -2.02 and the 95% confidence interval for the estimated
population mean difference is between -4.96 and 0.93. A paired sample test showed that the
difference between conditions was non-significant (t= -1.62, df= 7, p= 0.149, 2-tailed).
2
2.5
3
3.5
4
4.5
5
5.5
Chromatic Achromatic
Mea
n d
isp
arit
y th
resh
old
(ar
c m
ins)
0
1
2
3
4
5
6
Mean Disparity Threshold
In Phase
Out Of Phase
Figure 3. Mean disparity threshold for in phase and out of phase. Figure 4 showed the
average disparity threshold for four separated orientations: 0 (M= 4.37, SD= 2.88), 30 (M=
3.74, SD= 2.17), 60 (M= 4.08, SD= 2.45) and 90 (M= 2.23, SD= 2.14) degrees. There was a
significant effect (ANOVA?) of the degree of orientation, F(3,21) = 3.207, p= 0.044. Then a
pairwise comparison was carried out to show the difference between each individual degree
of orientation. It indicated that there was no significant difference between 0, 30, 60 and 90
degrees.
Figure 4. Mean disparity threshold for four different degrees of orientation.
Figure 5 showed the average disparity threshold for four individual phase variables: 0 degree