Comparison of Simultaneous Measurement While Viewing Real

Objects and 3D Video Clips

Tomoki Shiomi

, Keita Uemoto

, Takehito Kojima

, Satoshi Hasegawa

, Masako Omori

Hiromu Ishio

, Hiroki Takada

and Masaru Miyao

Nagoya Univ., Graduate Sch. of Inf. Sci., Chikusa-ku, Nagoya, Japan

Nagoya Bunri Univ. Inazawa, Aichi, Japan

Kobe Women's Univ., Suma-ku, Kobe, Japan

Fukuyama City Univ., Fukuyama, Hiroshima, Japan

Univ. of Fukui, Fukui, Japan

Keywords: Accommodation and Convergence, Simultaneous Measurement, Real Object, 3D Video Clips.

Abstract: The use of 3-dimensional images has been spreading rapidly in recent years such as in 3D films and 3D

televisions. However, the influence of stereoscopic vision on human visual function remains insufficiently

understood. The public has come to understand that lens accommodation and convergence are mismatched

while viewing 3D video clips, and this is the main reason for the visual fatigue caused by 3D. The aim in

this study is to compare the fixation distance of accommodation and convergence in viewing real objects

and 3D video clips. Real objects and 3D video clips perform the same movements. We measured

accommodation and convergence in viewing real objects and 3D video clips. From the result of this

experiment, we found that no discrepancy exists in viewing 3D video clips like real object. Therefore, we

argue that the symptoms in viewing stereoscopic vision may not be due to the discrepancy between lens

accommodation and convergence.

1 INTRODUCTION

Recently stereoscopic images have been used in

various ways. In spite of this increase in 3D products,

and the many studies that have been done on

stereoscopic vision, the influence of stereoscopic

vision on human visual function remains

insufficiently understood. Investigations of the

influence of stereoscopic vision on the human body

are essential in order to ensure the safety of viewing

virtual 3-dimensional objects People often report

symptoms such as eye fatigue and 3D sickness when

continuously viewing 3-dimensional images.

However, such problems are unreported with so-

called natural vision. One of the reasons often given

for these symptoms is that lens accommodation and

convergence (Fig.1) are inconsistent during the

viewing of 3D images (Fig.2) (Lambooij, 2009).

Accommodation is a reaction that occurs due to

the differences of refractive power by changing the

curvature of the lens with the action of the musculus

ciliaris of the eye along with the elasticity of the lens.

The result is that the retina focuses on an image of

the external world.

Figure 1: Lens accommodation and convergence.

Convergence is a movement where both eyes

rotate internally, functioning to concentrate the eyes

on one point in the front. The main method of

presenting 3-dimensional images is through the use

of the mechanism of this binocular vision.

We would like to argue that a discrepancy

between accommodation and convergence does not

542

Shiomi T., Uemoto K., Kojima T., Hasegawa S., Omori M., Ishio H., Takada H. and Miyao M..

Comparison of Simultaneous Measurement While Viewing Real Objects and 3D Video Clips.

DOI: 10.5220/0004278405420545

In Proceedings of the International Conference on Computer Graphics Theory and Applications and International Conference on Information

Visualization Theory and Applications (IVAPP-2013), pages 542-545

ISBN: 978-989-8565-46-4

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

exist, and we have already obtained results that

indicate the inconsistency between accommodation

and convergence does not occur in our previous

study (Miyao, 1996). In this present study, we

performed a more detailed investigations about

nonexistance of this discrepancy.

Figure 2: Discrepancy between accommodation and

convergence.

2 METHOD

We used an original machine developed by

combining WAM-5500

and EMR-9

to perform

the simultaneous measurements of accommodation

and convergence.

Subjects gazed in binocular vision at a real object

in natural vision (a Rubik’s cube) and then at a

virtual object of 3D video clips presented in front of

them. We measured their lens accommodation and

convergence (Fig.3). The objects viewed by the

subjects in natural and stereoscopic vision showed

exactly the same motion, and there were three kinds

of movements of these objects (Fig.4).

(1) The objects of natural and stereoscopic vision

moved forward and backward at a range from 0.5 to

1m with a cycle of 10 seconds. It was repeated four

cycles per single measurement.

(2) The second movement was the same motion as in

movement one, but the time of a single cycle of

movement was 2.5 seconds.

(3) The object in this movement approached the

subject. Initially, the position of the object was 1m

from the subject. The object moved forward to the

subject and stopped at the position of 1D, 1.5D, 2D

for each 10 seconds (D represents diopter).

A“diopter” is the refractive index of the eye lens,

which is an index of accommodation power. It

would be as follows 0D stands for infinity, 1D

stands for 1 m, and 2D stands for 0.5m.

Figure 3: The overview of this experiment process.

Figure 4: The movement of the object in natural and

stereoscopic vision. (1) First was a cycle of 10 seconds.

(2) Second was a cycle of 2.5 seconds. (3) Third was step

motion, the object stopped at 1D, 1.5D, and 2D for each

10seconds.

The measurements of the objects in both natural

and stereoscopic vision were taken three times per

one movement. The illuminance in this experiment

was 103 lx.

3 RESULT

The measurements for all subjects showed roughly

similar tendencies. Figures 5-8 shows the results of

movement 1 and 2, which is the moving object in

both natural and stereoscopic vision with a cycle of

10 seconds or 2.5 seconds.

In all figures, “accommodation” stands for the

focal length of lens accommodation, while

ComparisonofSimultaneousMeasurementWhileViewingRealObjectsand3DVideoClips

543

“convergence” stands for the convergent focal

length, and “object” stands for the location of the

real object in natural vision or the position of virtual

image in stereoscopic vision.

In Fig.5 and 6, these figures showed that the

accommodation and convergence of subjects

changed in agreement. The change in the diopter

value occurred within a cycle of about ten seconds.

The values of accommodation in both figures were

1.6D when object at the point of 2D and these values

were 1D when object at the point of 1D. The value

of convergence almost agreed with the position of

the object.

Figure 5: The result of natural vision (a cycle of 10

seconds).

Figure 6: The result of stereoscopic vision (a cycle of 10

seconds).

Figure 7: The result of natural vision (a cycle of 2.5

seconds).

In the case of movement 2 and 3, lens

accommodation and convergence almost agreed with

the position of the virtual images though the focal

point of accommodation often shifted to the far point

slightly.

Figure 8: The result of stereoscopic vision (a cycle of 2.5

seconds).

4 DISCUSSION

According to Hoffman et al. and Ukai and Howarth

(Ukai and Howarth, 2008), lens accommodation in

viewing 3D images would be fixed at the position of

the display. They have reported that an

accommodation-vergence mismatch can create

problems such as eyestrain and visual discomfort

due to the synergy between accommodation and

convergence. However, our experiment found no

mismatch between accommodation and convergence.

In our previous study, we also reported the results of

simultaneous measurement of lens accommodation

and convergence while subjects viewed objects in

stereoscopic vision, and the inconsistency between

accommodation and convergence did not occur

(Hori et al., 2011). This study simultaneously

measured accommodation and convergence in

viewing 3D video clips of three movements, and the

discrepancy was unconfirmed as in viewing real

object. Therefore, we found that subjects watching

3D so not show any discrepancy between

accommodation and convergence.

Subjects should be seeing blurred images if lens

accommodation focuses on the virtual image

position while a stereoscopic image project outwards.

Subjects focusing on a nearer position rather than

the display may be experiencing the condition in

which humans look at a position beyond the farthest

point of the object as in myopia.

Smith showed that the relationship between the

refractive error and visual acuity is linear (Smith,

IVAPP2013-InternationalConferenceonInformationVisualizationTheoryandApplications

544

1991). The visual acuity of subjects in Smith’s

experiment did not decrease much. Therefore, the

distance from an emerging object in our experiment

may not have been a problem and was correctly

viewed by subjects.

Meanwhile, Patterson reported that there should

be a problem in only a near-eye display and that the

accommodation-vergence mismatch likely would

not occur under most stereoscopic display viewing

conditions because of the depth of field (Patterson,

2009). Patterson (2009) and Wang and Ciuffreda

(2006) found that the depth of field was large, and

they stated that the average total depth of focus was

on the order of 1.0 diopter (Wang anda Ciuffreda,

2006). Based on this value, the range of total depth

of field would be from a distance of about 0.1m in

front of a fixed point to about 0.17m behind the

fixed point of 0.5m. For a fixed distance of 1 m, the

total depth of field would be from a distance of

about 0.33m in front of the point to about 1.0 m

behind the point. For a fixed distance of 2 m, the

total depth of field would be from about 1m in front

of the point to an infinite distance behind the fixed

point.

They also reported that the depth of field was

affected in various ways by the pupil diameter and

resolution.

Some researchers found that pupil diameter will

be slightly over 6 mm for a luminance level of

0.03cd/m2 and near to 2 mm for a luminance level

of 300cd/m2. For each millimeter of decrease in

pupil size, the depth of field increases by about 0.12

diopters (Patterson, 2009).

Therefore, the value of accommodation can be in

the range of the depth of field in our experiment.

In the future research, we plan further

investigations concerning the influence of age, pupil

diameter, the illuminance of the experimental

environment, and the luminance of visual targets.

5 CONCLUSIONS

In this experiment, we simultaneously measured

accommodation and convergence of subjects

viewing real object and 3D video clips that showed

exactly the same motion as in real objects. We did

not confirm the existence of discrepancy between

lens accommodation and convergence. Therefore,

we believe it is inconclusive that symptoms such as

eye fatigue and 3D sickness are not caused by this

discrepancy but other factors. We plan to perform

further investigations and studies of other variables.

REFERENCES

M. Lambooij, W. IJsselsteijn, M. Fortuin, and I.

Heynderickx, 2009. Visual Discomfort and Visual

Fatigue of Stereoscopic Displays: A Review, J.

Imaging Sci. Technol. Vol. 53, 03.

M. Miyao, et al., 1996. Visual accommodation and subject

performance during a stereographic object task using

liquid crystal shutters. Ergonomics, Vol.39, 11, 1294-

1309.

K. Ukai, P. A. Howarth, 2008. Visual fatigue caused by

viewing stereoscopic motion images: background,

theories, and observation, Displays 29, 2, 106-116.

H. Hori, T. Shiomi, et al., 2011. Comparison of

accommodation and convergence by simultaneous

measurements during 2D and 3D vision gaze. FORMA

Specieal Issue.

G. Smith, 1991. Relation between Spherical Refractive

Error and Visual Acuity, Optometry Vision Science,

Vol. 68, No.8, 591-598.

R. Patterson, 2009. Human factors of stereo displays: An

Update. Journal of SID, Vol.17, No.12, 987-996.

B. Wang and K. J. Ciuffreda, 2006. Depth of focus of the

human eye: Theory and clinical applications. Surv.

Ophthal. 51, 75.

ComparisonofSimultaneousMeasurementWhileViewingRealObjectsand3DVideoClips

545