Received August 16, 2017, accepted September 23, 2017, date of publication October 11, 2017,

date of current version November 14, 2017.

Digital Object Identifier 10.1109/ACCESS.2017.2761538

Depth Perception Assessment for Stereoscopic

3D Displays Using Layered Random
Dot Stereogram
MIN-KOO KANG , (Graduate Student Member, IEEE), AND SUNG-KYU KIM
Imaging Media Research Center, Korea Institute of Science and Technology (KIST), Seongbuk-gu, Seoul 02792, South Korea
Corresponding author: Sung-Kyu Kim (kkk@kist.re.kr)
This work was supported by the Cross-Ministry Giga KOREA Project through the Korean Government (MSIT) under Grant GK17D0200
(Development of Super Multi-View Display Providing Real-Time Interaction).

ABSTRACT Determining the features of human binocular vision according to given disparity information
has gained significant interest with regard to 3-D-relevant fields. However, subjectively distinguishing
the features is internally ambiguous owing to personal factors and external susceptibility due to viewing-
environmental factors. Moreover, high cost and low reproducibility of the subjective assessment method
restricts its usability in numerous practical applications. To alleviate these problems, this paper presents
a novel visual stimulus, namely layered random dot stereogram (LRDS), and hierarchical paradigm for
vote ratings. Based on the guide depth information in LRDS, viewers could recognize considerably obvious
feature points, and our two-step rating strategy enhanced the reliability and efficiency of the viewing tests.
The experiments prove that our approach successfully reflects the effects of not only personal factors but
also display factors on depth perception. Our approach would be beneficial to the 3-D fields of human factor,
viewer-interactive systems, and product performance evaluation.

INDEX TERMS Depth perception, RDS, VQA, stereo acuity, fusion limit, visual discomfort.

I. INTRODUCTION this limit was referred to as ‘‘zone of clear single binocular

Owing TO recent advancements in three-dimensional (3D) vision.’’ In [14], correction responses of the accommodation
technology, explicit factors causing visual discomfort (VD) and vergence mechanisms became unstable and oscillated at
have been alleviated. However, the VD elicited by the ver- the angular disparity of 2.6◦ , and the viewers experienced
gence accommodation conflict (VAC) is still inevitable in ‘‘diplopia.’’
stereoscopic 3D (S3D) displays [1]. Thus, measuring fea- To measure perceptual evaluation of the above FPs,
tures on depth perception and the level of VD has gained the subjective assessment methods (e.g. psychophysical scal-
significant interest of researchers in relevant fields [2], [3]. ing paradigms) are used, where viewers are asked to evaluate
Also, the approach to enhance final viewers’ quality of expe- and report their sensation of depth percepts for presented
rience (QoE) [4] based on perceptual depth quality [5], [6] 3D stimuli [2]. However, the conventional approach encoun-
has widely increased in various 3D applications. ters several technical challenges in various practical appli-
Well-known feature points (FPs) on depth perception cations. First, it is internally ambiguous since non-expert
concerning human binocular vision (HBV) are as follows. viewers have different standards for each FP owing to per-
In [7]–[9], the smallest angular disparity perceived as 3D was sonal experience, bias, preference, or misinterpretation. Also,
referred to as ‘‘stereo acuity,’’ and its range was known for the depth perception is susceptible to the extensive list of
1.8’’–50’’ (seconds of arc). In [1], [9], and [10], viewers com- factors concerning variants of S3D content, viewing envi-
fortably observed angular disparities less than 1.0◦ as comfort ronment, and hardware setup. Consequently, practical mea-
depth, and the limit of 1.0◦ was referred to as ‘‘Percival’s zone surements should differ from the statistical agreements of
of comfort.’’ In contrast, a disparity of more than 1.0◦ was the above reports. Second, it is time-consuming and difficult
known to induce severe VD [3]. With respect to binocular to reproduce the same results of the advanced experiments
fusion, [12], [13] reported that the limit where the correction owing to the lack of open-source tools for stimulus generation
responses for VAC are achievable is approximately 2.0◦ , and and evaluation.
2169-3536 2017 IEEE. Translations and content mining are permitted for academic research only.
VOLUME 5, 2017 Personal use is also permitted, but republication/redistribution requires IEEE permission. 22855
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
 M.-K. Kang, S.-K. KIM: Depth Perception Assessment for Stereoscopic 3-D Displays Using LRDS
FIGURE 1. Overall concept of proposed approach. (a) Five significant
feature points (FPs) concerning depth perception. (b) Self-depth
perception assessment scenario using proposed visual stimulus and
presentation tool.
This study presents a new methodology for subjective
depth perception assessment (DPA). Conventionally, distinct
DPA experiments for finding each HBV FP shown in Fig. 1(a)
have been conducted. However, the proposed method reliably
and cost-efficiently measures the five HBV FPs at once while
overcoming the above technical challenges for its practical
use in various S3D video applications. To this end, new types
of test stimulus and vote rating paradigm were developed. The
overview of the proposed method is depicted in Fig. 1(b). For
the test stimulus, we designed a new functional 3D image
called layered random dot stereogram (LRDS). It aids to
reduce the ambiguity of depth percepts with auxiliary min-
max depth information. For the stimulus presentation and
vote rating, a concept of hierarchical rating paradigm is intro-
duced. It assigns two distinctive tasks to viewers, and the
correct response for the first task is required to viewers to
proceed the second task. This strategy aids viewers to focus
on the cognitive process for accurate depth perception and
enables outliers to be removed efficiently as saving the exe-
cution time for tests. It is inspired by the oddball paradigm of
event-related potential (ERP) research [15], [16], a substan-
tially widely used method for studying the cognitive process
in neuroscience. For the computation of the five FPs, three
different operations are applied to distributions of vote ratings
after a screening of bad vote ratings based on an additional
integrity check. This results in boosting the reliability of the
FPs of non-expert viewers.
The concept of the proposed approach was initially intro-
duced in the previous study [17] for the application of
calibrating personal binocular vision characteristics. In this
paper, we present more generalized methodology and its
open-source tools for generating the proposed LRDS stim-
ulus with various S3D formats and automatic processing
of the viewing test for non-expert users. Also, we demon-
strate two explorative experiments as applications of the pro-
posed method and tools. The experimental results show the
22856
FIGURE 2. Layered random dot stereogram (LRDS). (a) Locations of each
layer in 3D space. (b) Mask for shape layers where the direction of mid
layer is right. (c) Practical LRDS sample. (d)(e)(f) Conceptual images in the
human brain according to conditions of personal depth perception.
feasibility of the proposed method and tools as the reference
source in various S3D video research fields.
II. PROPOSED METHOD
A. VISUAL STIMULUS: LAYERED RANDOM DOT
STEREOGRAM (LRDS)
In the field of DPA, RDS is often used to investigate mech-
anisms and processes of stereopsis because it only contains
binocular disparity information so that viewers cannot recog-
nize any meaningful information without the correct binocu-
lar fusion [18]. Based on the principle of RDS, the proposed
LRDS consists of three RDS layers (e.g. Fig. 2(a)), and each
layer has a unique function. The rear layer is located at
viewing distances, and contributes to assessing stereo acuity.
The front layer is located at depth planes extremely close
to the viewers where clear single binocular vision should
not be formed; it contributes to determining the limits of
binocular fusion. The mid layer is located at one of depth
planes between layers at the front and rear, and its location
becomes the target depth that is to be evaluated.
In addition, LRDS has shape variants. The shapes of each
layer are shown in Fig. 2(b): a large circle, a modified Maltese
cross, and caps that provide the holes of Maltese crosses for
the layers at the rear, mid, and front, respectively. Notice that
the small center circle is connected with one of four triangles
in the mid layer. In this manner, we produced four directional
shape variants; however, users could adjust the number of
directions by changing the rotation angle. This adjustment
VOLUME 5, 2017
M.-K. Kang, S.-K. KIM: Depth Perception Assessment for Stereoscopic 3-D Displays Using LRDS
would affect the integrity and execution time of the proposed
approach.
With depth and shape variants, we could then further
evidently investigate HBV FPs on depth perception. For
instance, if a test depth is considerably small compared
to viewers’ stereo acuity, subsequently, the viewers could
not recognize any shape information of the mid layer as
in Fig. 2(d). Further, if a test depth exceeds viewers’ binocular
fusion limit, the viewers might experience diplopia, and the
blurred layers of the front and mid are attached together
concealing the direction information as shown in Fig. 2(f).
Otherwise, viewers could recognize the exact depth and direc-
tion of the mid layer as shown in Fig 2(e), and give correct
vote ratings.
FIGURE 3. Stimulus presentation and vote rating methods.
(a) Conventional double-stimulus-impairment scale (DSIS) method and its
five-grade ratings. (b) Flow chart of proposed method, keypad buttons for
hierarchical voting, and our qualitative three-grade scale ratings.
B. STIMULUS PRESENTATION: HIERARCHICAL
RATING PARADIGM
For the stimulus presentation, an unimpaired reference and
the same picture impaired are sequentially presented in
the general video quality assessment (VQA) (e.g., DSIS
method) [2]. Viewers are then asked to vote on the qual-
ity of impaired picture with the five-grade scale as shown
in Fig 3(a).
Based on the conventional DSIS method, we developed a
more sophisticated method for the stimulus presentation and
vote ratings using a hierarchical rating paradigm. Its flow
chart is depicted in Fig 3(b). First, a 2D image that has zero
disparity with the same RDS pattern is presented as the ref-
erence. It initializes the viewers’ ability to recognize a binoc-
ular disparity by equalizing vergence and accommodation
VOLUME 5, 2017
mechanisms critically linked with VD [1], [3]. Secondly,
one of the prepared LRDS stimuli is randomly selected and
presented to the viewer. Thereafter, two hierarchical vote
ratings are conducted as follows. In the first vote, the question
is ‘‘which triangle is connected with the centric circle in the
mid layer?’’ Subsequently, viewers should push one of the
four direction buttons corresponding to the correct direction.
If viewers fail to recognize the right direction, it is recom-
mended to press the question marked button. In the second
vote, viewers should push one of the lettered buttons corre-
sponding to the QoE grade for the presented LRDS. Since the
first vote semantically filters out the marginal grades of the
conventional five-grade scale, we only use three-scale grades
that describe the quality of depth perception concerning
3D effect and VD. Notice that if viewers provide a wrong
answer in the first vote, the second vote is omitted as a time
saving strategy.
C. VOTE INTEGRITY CHECK AND FEATURE
POINT COMPUTATION
In the conventional VQA method, the mean rating is used
for each distribution of ratings as the representative value
after screening of the observers by ascertaining whether the
distribution of ratings for each test is normal or not for better
analysis of the results. Similarly, the proposed DPA method
have methods for computing the representative values of the
five FPs and screening of undesired data. However, it con-
ducts these processes with a more sophisticated manner as
follows.
After evaluating all trials, we have three distribu-
tions of vote ratings corresponding to each quality grade:
‘‘L,’’ ‘‘C,’’ and ‘‘D.’’ However, the raw distributions could
have undesirable votes caused by changes of personal con-
dition, errors in button pressing, etc. Also, it is necessary
to consider doubt guesses or cheated votes owing to the
insincere attitude of viewers. Thus, it is critical to define
the substantial representative and reliable FPs from the raw
distributions of vote ratings.
To this end, raw distributions of vote ratings are defined as
in (1) after sorting the stimulus levels with ascending order:
n o
j j
Ui = υik |1 ≤ κ ≤jcount and j ∈ { L, C, D} (1)
where i is a viewer index, j is an index for the quality
classes, and k is an index for the stimulus counts in each
quality class. Subsequently, the five HBV FPs are derived
as follows. For the representative levels of the three quality
grades, the median values of each distribution are selected.
If the depth step-sizes of stimuli are sufficiently small, and
the number of directional shape-variants of LRDS is suf-
ficiently large, three FPs become more reliable. For stereo
acuity and fusion limit, the minimum level of UiL and
the maximum level of UiD could be selected, respectively.
However, the marginal levels have a high risk of being out-
liers after the sorting process. Thus, an additional integrity
check is required. In our LRDS stimuli, four consecutive
22857
 M.-K. Kang, S.-K. KIM: Depth Perception Assessment for Stereoscopic 3-D Displays Using LRDS

FIGURE 4. Example of LRDS stimulus generation. (a) Shape mask samples that have four shape variants and t depth variants, where intensities of each
layer indicate disparities of RDS pattern. (b) LRDS samples converted to multi-view S3D image format.

levels have the same depth since they are directional variants. single binocular vision. Finally, intensities corresponding to
Therefore, more than two stimulus levels having the same the test disparity levels should be allocated to the pixels in the
depth should belong to the same distributions. From this mid layer.
principle, the additional integrity check for final stereo acuity Users can change the shape and the numbers of direction
and fusion limit are performed with Algorithm 1. variant, but the changes may affect the accuracy and exe-
cution time of viewing tests since it determines the number
Algorithm 1 Integrity Check of the Raw Vote Ratings of stimuli. The number of test stimuli is calculated by the
1: procedure STEREO ACUITY (UiL ) F An integrity check number of test disparity levels multiplied by the number of
for stereo acuity shape variants. In Fig. 4, the number of stimulus levels is four
2: SAi ← minUiL F SAnewi is the current times greater than the practical number of the test disparity
stereo acuity value levels since we used four directional variants in the stimuli.
3: SAnew
i ← min(UiL − {SAi }) F SAnew i is a temporal Fig. 5 shows practical LRDS capture images in a glasses-
stereo acuity value free S3D display (side-by-side type), where two capture
new ← U
4: UiL iL methods were used for the comparison: screen capture
5: While (SAnew i -SAi )≥ 4 or {(SAi
new + 3)%4 − (SA + 3)
i (digital capture) and camera capture (optical capture). Since
%4} < 0 do the left and right images of S3D contents are superimposed
6: UiLnew ← U new − {SA } F SAi is regarded as an
iL i in S3D displays, the screen capture images showed slightly
outliner blurred shapes of the Maltese cross in accordance with dis-
7: SAi ← minUiL new F A stereo acuity is newly parity increase. However, images captured by the camera
updated intending a virtual human eye revealed that there are no
8: SAnew
i ← min(U new -{SA })
iL i noticeable differences except for the random dot distribu-
final
9: SAi ← SAi tion. This implies that any shape and depth information is
final concealed in the proposed LRDS without correct binocular
10: return SAi F The final stereo acuity
value SAi
final depth perception, and pure cognitive tests concerning to depth
perception are possible for given disparity levels.

III. STIMULUS GENERATION AND VIEWING TEST B. VIEWING TEST SETUP

A. TEST LRDS GENERATION
In order to generate LRDS stimuli, shape masks should be
prepared in advance. Then, the proposed LRDS generation
tool automatically converts the shape mask input to the LRDS
output according to a target S3D image format: side by side,
top and bottom, or multi-view. Fig. 4 shows examples of
shape masks and those corresponding LRDS stimuli. Similar
to the conventional depth map, the intensities of each layer
in a shape mask indicate the disparities for RDS pattern shift.
The pixel intensity in the rear layer should be ‘‘0’’ for zero
disparity, and the pixel intensity in the front layer should be
sufficiently large to prevent recognizing the front layer as
For a viewing test, a set of shape masks and correspond-
ing LRDS stimuli, a target S3D display, a keypad for a
button press, and a subject should be prepared as depicted
in Fig. 6. According to the flowchart in Fig. 3, whole
processes are conducted automatically corresponding to the
viewer response for every trial. As a result, subjects can
evaluate their own FPs of depth perception by using the
proposed tool for viewing tests. Although the proposed
approach is developed for self-evaluation of non-expert
viewers, an additional 2D display can be connected for
the behavioral monitoring of subjects. In this case, the
2D display should be presented to the observers only, and

22858 VOLUME 5, 2017

M.-K. Kang, S.-K. KIM: Depth Perception Assessment for Stereoscopic 3-D Displays Using LRDS
FIGURE 5. Test LRDS sample comparison. Screen captured and camera captured LRDS samples were compared. Left and right images of LRDS samples
are superimposed when it is screen captured, but one of left and right images is shown when it is captured by a camera.
FIGURE 6. Overview of viewing test. (Left) proposed tool for viewing
tests. (Right) example of subject self-evaluation.
step-by-step information such as the trial number, the correct
direction for the current stimulus, and the viewer answer can
be inspected in detail from the control panel of the proposed
tool. If all trials are viewed, then the tool also prints out the
log of the experiment, and computes the five FPs based on
the proposed integrity check algorithm.
IV. APPLICATIONS AND EXPERIMENTAL RESULTS
The proposed method and tools would be beneficial to various
S3D video industries. For instance, in the case of S3D content
production, it can aid stereographers to plan a better depth
budget for comfort S3D viewing while providing a DB of
perceptual tolerances for the use of excessive binocular dis-
parity. In the case of manufacturing of S3D devices such as
S3D displays, it can be used for performance evaluation of
their products since the range of valid depth expression or
the space of comfort zone could be a quantitative quality
measure. In the case of S3D video broadcasting services,
it can give reliable standards to set guidelines for the safe
S3D content viewing. Also, it can be cited as a reference tool
in academic fields such as 3D human factor research or med-
ical research. To show the feasibility of the proposed method,
we demonstrate two explorative experiments in this section
as follows.
A. APPLICATION 1: PERSONAL HBV FP INVESTIGATION
The first experiment was designed to investigate personal
HBV FPs elicited by inherent characteristics of the viewers.
VOLUME 5, 2017
FIGURE 7. Schematic plot for experimental setup and test levels of depth.
To this end, all the other factors, related to the 3D content and
viewing environment [3]–[7], that can affect the HBV FPs
were restrained. For this test, four non-expert subjects, aged
between 25 and 34 years, attended, and we generated 128 lev-
els of LRDS. Since four consecutive LRDS levels, with dif-
ferent directions, have the same disparity level, 32 levels of
disparity were investigated in practice. The same glasses-free
S3D display was used in the same viewing environment for
the test. The schematic plot for this experiment is depicted
in Fig. 7. The accommodation distance was maintained as
623 mm, and the mid layers of test LRDS stimuli varied
between 619 mm and 313 mm in terms of depth from the sub-
jects. The first six depth levels were designed to have the step
size of 1 mm for stereo acuity, and the others were intended
to have the step size of 10 mm. The angular disparity can
be calculated from the given accommodation distance, depth
levels, and the interpupillary distance where we assumed it as
the average length (65 mm).
Fig. 8 illustrates the experimental results. The left graph
shows the raw distributions of the stimulus level belonging to
each quality grade, rated in the second vote of the hierarchical
22859
 M.-K. Kang, S.-K. KIM: Depth Perception Assessment for Stereoscopic 3-D Displays Using LRDS

FIGURE 8. Experimental results for the first application. (a) Distributions of the raw votes to each quality grade for four different viewers. (b) Human
binocular vision feature points after outlier removal via proposed integrity check.

FIGURE 9. Distributions of raw votes for each quality grade. Comparison of depth expression of three different types of 3D displays.

rating paradigm. Although the correct depth perception is
verified in the first vote, the wrong quality evaluation could
be made for numerous reasons. In this graph, we indicated
the unreliable votes with a dotted circle, determined by the
proposed integrity check process. The right graph shows
the final angular disparities corresponding to the five HBV
FPs after the unreliable vote removal, where the dotted lines
indicate the two meaningful FPs, the comfort limit and the
fusion limit, commonly reported in the literature.
In a comprehensive sense, all subjects showed similar
trends. This result indicates that the proposed approach is
considerably free from the internal ambiguity although the
subjects are not experts in DPA. Also, the experimental results
confirmed the fact that perceptual HBV FPs slightly differ
from the statistical HBV FPs [1], [3], [7], [9], [10], [12]–[14]
owing to numerous variants of personal factors. For instance,
although the limit of Percival’s zone of comfort [1] (1.0◦ )
belong the dynamic range of ‘‘C’’ grade, the representative
disparity slightly oscillates near 0.68◦ across all subjects.
Similarly, although the limit of the zone of clear single binoc-
ular vision is approximately 2.0◦ [12], [13], the binocular
fusion limit in our test is about 2.72◦ on average. In the
individual results, V2 demonstrated the widest dynamic range
for single binocular fusion, which indicates that V2 is the
most sensitive to disparity changes, and the most tolerant of
excessive depths. In contrast, V3 demonstrated the narrowest
dynamic range for single binocular fusion. This result indi-
cates that V3 is less sensitive to disparity changes, and needs
for special caution of viewing excessive depths.
Additionally, we can analyze each subject’s certainty
(or could be consistency) of the quality evaluation for per-
ceived depth by counting the unreliable vote. In this sense,
V2 registered only three unreliable votes, and can be regarded
as the most-favored tester. On the contrary, V3 registered
twelve unreliable votes, and thus can be considered as an
inappropriate subject for this kind of DPA tests. In this
manner, the proposed method and tools can be exploited for
screening inappropriate viewers relevant to the judgment.
B. APPLICATION 2: PERFORMANCE COMPARISON
OF 3D DISPLAYS
The second experiment was designed to compare the per-
formance of 3D displays concerning to the ability of depth
expression. To this end, three expert-subjects participated,
and the representative disparity values at the five FPs derived
by three different glasses-free 3D displays were compared,
where all the other factors that can affect the results were also
restrained. The only variants in this experiment are the display
factors illustrated in Table 1 such as accommodation distance
and interocular crosstalk. Therefore, it is regarded that the
differences at the five FPs for the same subject are elicited
by the properties of the test displays. The experimental

22860 VOLUME 5, 2017

M.-K. Kang, S.-K. KIM: Depth Perception Assessment for Stereoscopic 3-D Displays Using LRDS
TABLE 1. Specifications of test displays for the experiment of second application.
setup was similar to the first experiment, but test depth range
and step sizes were distinctively designed according to the
specifications of the test displays, where step size A and B
were for stereo acuity and the other FPs, respectively.
The experimental results are illustrated in Fig. 9 and
Table 2. Similar to the right graph of Fig. 8, Fig 9 shows
the raw distributions of the stimulus level belonging to each
quality grade. Since the interests of this experiment are char-
acteristics of the test displays, unreliable votes were not
highlighted in Fig. 9. Instead, the exact angular disparities for
each FPs are summarized in Table 2 after the unreliable vote
removal.
TABLE 2. Results of HBV features in the second application.
The range of single binocular vision, stereo acuity to
fusion limit, could be referred as an objective performance
metric for 3D displays. In this perspective, display A
demonstrated the maximum narrow range (perception range:
0.040◦ –1.324◦ /test range: 0.017◦ –1.712◦ ) on average.
We assume that this is deduced from the conventional dis-
advantage of multi-view displays. Although multi-view dis-
plays can provide motion parallax even for still images, those
have a disadvantage of surplus blur caused by adjacent-view
crosstalk, compared to stereoscopic displays. Especially for
display B and C, subjects V6 and V7 correctly perceived the
min-max test depth thus FP1 and FP5 were not evidently
compared, and certain FP1s and FP2s demonstrated the same
angular disparities. However, subject V5, the most intolerant
of excessive depths, recognized that display B apparently
provides higher fusion limit compared to that of display C.
VOLUME 5, 2017
This is consistent with the prior research [3], [7], where
the fusion limit was elevated at low spatial frequencies, and
display C had a low spatial frequency induced by its high inte-
rocular crosstalk. Also, FP2, FP3, and FP4 could be regarded
as quantitative quality standards for 3D displays. In this
perspective, display B demonstrated the highest disparities
at FP3 even for subjects V6 and V7. This result indicates that
display B has the widest zone of comfort.
V. CONCLUSION
In conclusion, we presented a novel method and tools for sub-
jective depth perception assessment. We especially focused
on the fact that viewers could provide wrong votes owing
to the following factors: errors in ratings, guessed votes,
biased judgment, changes in concentration, etc. To this end,
stimuli eliciting inaccurate depth perception are filtered prior
to their quality evaluation, and additional integrity for vote
ratings is checked using the proposed LRDS stimulus. The
cost efficiency is enhanced by our presentation paradigm of
hierarchical vote ratings. The whole processes, a stimulus
presentation to the computation of HBV FPs, are conducted
automatically. Thus, non-expert viewers can assess their own
FPs of depth perception without external aid. To show the
usability for various applications, we also demonstrated two
explorative experiments. Being open-source codes, tools for
LRDS generation and stimulus presentation are available
on the first author’s homepage. These would be beneficial
to various S3D video relevant fields such as 3D human
factor research, personalized 3D content service, 3D hard-
ware performance comparison, developing guidelines for safe
3D viewing with only simple manipulations of particular
key-parameters according to the objectives of the target
applications.
REFERENCES
[1] D. M. Hoffman, A. R. Girshick, K. Akeley, and M. S. Banks, ‘‘Vergence–
accommodation conflicts hinder visual performance and cause visual
fatigue,’’ J. Vis., vol. 8, no. 3, p. 33, Mar. 2008.
[2] L. M. J. Meesters, W. A. IJsselsteijn, and P. J. H. Seuntiens, ‘‘A survey
of perceptual evaluations and requirements of three-dimensional TV,’’
IEEE Trans. Circuits Syst. Video Technol., vol. 14, no. 3, pp. 381–391,
Mar. 2004.
[3] M. Lambooij, M. Fortuin, I. Heynderickx, and W. IJsselsteijn, ‘‘Visual
discomfort and visual fatigue of stereoscopic displays: A review,’’ J. Imag.
Sci. Technol., vol. 53, no. 3, pp. 30201-1–30201-14, 2009.
[4] M.-K. Kang, H. Cho, H.-M. Park, S. C. Jun, and K.-J. Yoon, ‘‘A well-
ness platform for stereoscopic 3D video systems using EEG-based visual
discomfort evaluation technology,’’ Appl. Ergon., vol. 62, pp. 158–167,
Jul. 2017.
22861
 M.-K. Kang, S.-K. KIM: Depth Perception Assessment for Stereoscopic 3-D Displays Using LRDS
[5] F. L. Kooi and A. Toet, ‘‘Visual comfort of binocular and 3D displays,’’
Displays, vol. 25, nos. 2–3, pp. 99–108, Aug. 2004.
[6] J. Wang, S. Wang, K. Ma, and Z. Wang, ‘‘Perceptual depth quality in
distorted stereoscopic images,’’ IEEE Trans. Image Process., vol. 26, no. 3,
pp. 1202–1215, Mar. 2017.
[7] Y.-Y. Yeh and L. D. Silverstein, ‘‘Limits of fusion and depth judgment
in stereoscopic color displays,’’ Hum. Factors, J. Hum. Factors Ergonom.
Soc., vol. 32, no. 1, pp. 45–60, 1990.
[8] D. B. Diner and D. H. Fender, Human Engineering in Stereoscopic Viewing
Devices. New York, NY, USA: Plenum Press, 1993.
[9] J. Kim et al., ‘‘Glasses-free randot stereotest,’’ J. Biomed. Opt., vol. 20,
no. 6, p. 065004, Jun. 2015.
[10] A. S. Percival, The Prescribing of Spectacles, 3rd ed. Bristol, U.K.: J.
Wright & Sons, 1928.
[11] M. Wöpking, ‘‘Viewing comfort with stereoscopic pictures: An exper-
imental study on the subjective effects of disparity magnitude and
depth of focus,’’ J. Soc. Inf. Display, vol. 3, no. 3, pp. 101–103,
Dec. 1995.
[12] C. Sheard, ‘‘Zones of ocular comfort,’’ Optometry Vis. Sci., vol. 7, no. 1,
pp. 9–25, Jan. 1930.
[13] I. P. Howard, Seeing in Depth, Basic Mechanisms, vol. 1. Toronto, ON,
Canada: Univ. Toronto Press, 2002.
[14] K. Ukai and Y. Kato, ‘‘The use of video refraction to measure the dynamic
properties of the near triad in observers of a 3-D display,’’ Ophthalmic
Physiol. Opt., vol. 22, no. 5, pp. 385–388, Sep. 2002.
[15] T. C. Handy, Event-Related Potentials: A Methods Handbook. London,
U.K.: MIT press, 2005.
[16] H.-O. Li, J. Seo, K. Kham, and S. Lee, ‘‘Measurement of 3D visual fatigue
using event-related potential (ERP): 3D oddball paradigm,’’ in Proc. 3DTV
Conf., Istanbul, Turkey, May 2008, pp. 213–216.
[17] M.-K. Kang and S.-K. Kim, ‘‘Personal binocular vision calibration using
layered random dot stereogram,’’ in Proc. APSIPA ASC, Jeju, South Korea,
Dec. 2016, pp. 1–4.
[18] B. Julesz, ‘‘Binocular depth perception of computer-generated patterns,’’
Bell Labs Tech. J., vol. 39, no. 5, pp. 1125–1162, Sep. 1960.
22862
MIN-KOO KANG (GS’10) was born in Ulsan
City, South Korea, in 1984. He received the
B.S. degree in electronic engineering from Inha
University, South Korea, in 2008, the M.S.
and Ph.D. degrees in electrical engineering
from the Gwangju Institute of Science and
Technology, South Korea, in 2010 and 2015,
respectively.
Since 2015, he has been a Post-Doctor with the
Center of Imaging Media Research, Korea Insti-
tute of Science and Technology, Seoul, South Korea. His current research
interests include super-multiview video and high-density multiview video
signal processing, 3-D human factor, and viewer-interactive video systems.
SUNG-KYU KIM received the B.S., M.S., and
Ph.D. degrees from the Quantum Optics Group
of Physics, Korea University, Seoul, South Korea,
in 1989, 1991, and 2000, respectively. Then, he
spent 2 years as an Invited Research Scientist,
TAO, Japan. In 2001, he was appointed as a
Research Scientist with the Korea Institute of Sci-
ence and Technology (KIST).
He is currently the Principal Research Scientist
with KIST. As a Senior Researcher, he has led the
Visual Media Research and Development Team, KIST, and succeeded in
developing original technologies allowing viewers to enjoy 3-D displays
without wearing special glasses and compromising resolution. His cur-
rent research interests also include the next-generation 3-D displays, such
as super-multiview video displays and multi-focus AR/VR head-mounted
displays.
VOLUME 5, 2017