Correct blur and accommodation information is a reliable cue to depth ordering
Marina Zannoli , Rachel A. Albert , Abdullah Bulbul , Rahul Narain , James F. O’Brien , Martin Banks
1
1
1
2
Vision Science Graduate Group, University of California, Berkeley, USA
1
Department of Computer Science, University of California, Berkeley, USA
Experiment 1
Blur & Occlusion
Experiment 2
Methods
Occlusion edge blur used as cue to depth ordering (Marshall et. al).
Methods
D
1.2
near fix.
a.
1
1
corresponding author: marinazannoli@gmail.com
Motivation
2
is
e
r
xtu
e
t
h
hic
far fix.
W
:
k
Tas
a.
b.
?
t
n
o
r
f
n
i
t?
n
ro
f
n
si
i
e
tur
Viewing configuration in Experiment 2. Single-plane presentation: two flat
textures displayed in same focal plane (3.2, 2.6, 2.0 or 1.4 D), artificially rendered blur (1.2 D). Monocular presentation.
ex
t
h
:
k
s
Ta
0.9
0.9
0.9
0.8
0.8
0.8
0.7
0.6
0.5
0.4
0.3
0.2
3.2 D
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Single−plane
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0
Volumetric
1
1
0.9
0.9
0.8
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
4 mm
5 mm
6 mm
7 mm
p < .05
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.3 sec
0
3 sec
Presentation x Fixed plane
Proportion correct
- Improved performance with correct accommodation cues?
- Reduction in individual biases with correct accommodation cues?
0.8
0.1
subjects’ pupil sizes
Question
0.9
Simulated Pupil Size
Proportion correct
High-speed switchable lens (Love et. al). (a) Schematic of lens. Four
focal states. (b) Top view of the lens. Viewing frustum in green, horizontal lines show focal distances of four lens states.
0.9
Proportion correct
2.6 D
0.8
p < .05
1
Far (blurred edge)
Presentation x Duration
p < .05
1
0.9
0.7
0.6
0.5
0.4
0.3
0.2
0.8
0.6
0.5
0.4
0.3
0.2
0.1
0
0
Near
Far
Volumetric
0.6
0.5
0.4
0.3
0.2
0.1
Near (sharp edge)
Far (blurred edge)
0
0.7
0.6
0.5
0.4
0.3
0.2
0.1
3.2 D
2.6 D
2.0 D
1.4 D
0
4 mm
5 mm
6 mm
Fixed Plane
1
0.9
0.3 sec
3 sec
Single−plane
0.3 sec
3 sec
Volumetric
7 mm
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Near
Exp. 1
Far
Near
Exp. 2
Far
Near
Far
Marshall et al.
- Bias for perceiving blurred texture
closer independant of edge blur.
- Consistant bias over time.
- Reduction in individual biases with
correct accommodation cues.
- No effect of viewing distance on
depth ordering.
References
0.7
0.1
Near
Far
Single−plane
Near (sharp edge)
0.7
Discussion
Fixed Plane
Proportion correct
monitor
Proportion correct
0.9
Proportion correct
2.0 D
p < .05
1
Proportion correct
1
Proportion correct
1
ANOVA: Presentation * Duration * Fixed Plane * Simulated Pupil Size
1
Ferroelectric liquid-crystal polarization switch, b Birefringent lens
1
0
Duration
Simulated Pupil Size
subjects’ pupil sizes
1.4 D
Fixed Plane
0.1
Proportion blurred texture closer
FLC 2
BF 2
FLCa 1
BFb 1
Results
Proportion correct
Stimuli used in Experiment 1. (a) Volumetric presentation: two flat textures
(chosen among four) displayed in separate focal planes (distance 1.2 D), blur
produced by accommodation. (b) Single-plane presentation: two flat textures
displayed in same focal plane, artificially rendered blur (1.2 D). Monocular
presentation.
Presentation
a
Results
Fixed Distance
Accommodation
polarizer
far fix.
ANOVA: Fixed Distance * Fixed Plane * Simulated Pupil Size
Bias: sharp boundary, blurred texture perceived in front.
Limitations:
- correct blur only if viewer accomodates to display screen.
- unrealistic blur.
b.
near fix.
b.
Ecological relationship between blur and occlusion. (a) Eye focused on far plane
seen through aperture. (b) Eye focused on near plane. Blur of occlusion boundary
identical to blur of near surface.
a.
ic
h
W
Love, Hoffman, Hands, Gao, Kirby, & Banks. (2009) High-speed switchable lens
enables the development of a volumetric stereoscopic display. Optics Express, 17, 15716-15725.
Marshall, Burbeck, Ariely, Rolland, &Martin. (1996) Occlusion edge blur: a cue
to relative visual depth. J. Opt. Soc. Am. A 13, 681-688.