What determines the reference framework for stereopsis? Andrew Glennerster, Martin Birch and Suzanne McKee* University Laboratory of Physiology, Oxford *Smith-Kettlewell Eye Research Institute, San Francisco
What determines the reference framework for stereopsis? Andrew Glennerster, Martin Birch and Suzanne McKee* University Laboratory of Physiology, Oxford.
Dec 20, 2015
Welcome message from author
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
What determines the reference framework for stereopsis?
Andrew Glennerster, Martin Birch
and Suzanne McKee*
University Laboratory of Physiology, Oxford
*Smith-Kettlewell Eye Research Institute, San Francisco
Viewing geometry
Depth in fronto-parallel coordinates
L R
Fixation plane
Depth relative to a local surface
L R
Mitchison and McKee (85, 87)Mitchison and Westheimer (84)
Reference plane
Depth perpendicular to a local surface
L R
Prediction
L R
Thresholdplanes,accordingto‘surface’hypothesis
Thresholdsshould all bethe same,according to‘fronto-parallel’hypothesis
Prediction
L R
Thresholdplanes
Interleaved psychometric functions
L R
Thresholdplanes
Stimulus details
Target column
4
Task• is central column in front or behind the plane of the grid?
• dot size 2 arcmin• pixel size 0.45 arcmin• anti-aliased• viewing distance 2.65 m• exposure 600 ms• grid disparity gradient +/- 0.1• slant about a vertical axis• modified Wheatstone stereoscope
Stimulus
Psychometric procedure• method of constant stimuli• 7 disparities per psychometric function (at least 1400 trials)• 3 psychometric functions interleaved …
Results: grid disparity gradient -0.1
Subject MDB
Grid disparity gradient +0.1
Subject MDB
Grid disparity gradient -0.1
Subject AG
Grid disparity gradient +0.1
Subject AG
Summary of thresholds
0
0.2
0.4
0.6
0.8
R cyc L R cyc L
Thr
esho
ld (
arcm
in) MDB 12% difference
p < 0.001(one-tailed t-test)
23% diff.p < 0.0001
Summary of thresholds
0
0.2
0.4
0.6
0.8
R cyc L R cyc L
Thr
esho
ld (
arcm
in) AG
15% differencep < 0.0112% diff.
p < 0.01
Summary
Thresholdplanes
• Thresholds for detecting relative disparity depend on the local surface slant
• Thresholds are lowest when the displacement is more orthogonal to the surface
Conclusion
• The visual system may represent depth in a surface-based coordinate frame ...
• … and hence be much less sensitive to rotations and translations of the head
Conclusion
• The visual system may represent depth in a surface-based coordinate frame ...
• … and hence be much less sensitive to rotations and translations of the head
Comparison with lateral displacement thresholds Key:
F-P grid, lateral judgement
Width of grid in the eye in whichthe target was displaced
(arcmin)
Target columnTarget column
Slanted grid (+0.1) depth judgementSlanted grid (-0.1) depth judgement
Subject MDB
Sensitivity to lateral displacement
Subject MDB
… and to disparity + lateral displacement
Subject MDB
… and to disparity + lateral displacement
Subject MDB
Grid gradient 0.1
Subject MDB Subject SPM
Grid gradient +0.1 and -0.1
Subject MDB
Local surface model
Subject MDB
+0.1 gradient -0.1 gradient
Interleaved slants
Subject MDB
+0.4 arcmin disparity -0.4 arcmin disparity
Effect of perceived slant
Subject MDB
Grid perceived as fronto-parallel:
Grid perceived as slanted:
Related Documents
