Content uploaded by Seong-Yoon Shin
Author content
All content in this area was uploaded by Seong-Yoon Shin on Oct 22, 2014
Content may be subject to copyright.
Bibliographic Info: J Intell Inform Syst 2013 December: 19(4): 11~20 11
Correlation between Car Accident and Car Color for
Intelligent Service
Seong-yoon Shin
Department of Computer Information Engineering,
Kunsan National University
(s3397220@kunsan.ac.kr)
Sangwon Lee
Division of Information and Electronic
Commerce (Institute of Convergence and
Creativity), Wonkwang University
(sangwonlee@wku.ac.kr)
․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․
In designing Intelligent Traffic Systems, it should be necessary to consider telecommunications,
appearance, environment, auxiliary functions, safety, and so on. Also, in choosing a car, a consumer
considers those properties. This paper tried to elucidate the fact that car color has a very significant
meaning for car safety when administrating intelligent traffic services and making car-purchasing
decision. We first studied on occurrence probability of car accident according to car color that has
something to do with car safety. Then, we studied on the concepts of advancing color and receding
color. Advancing color causes less accidents since the color looks closer than it actually is. And receding
color causes more accidents since the color looks farther than it actually is. And we classified car
colors into eight classes and assign their ranking to each class, considering the number of car accidents.
We tried to verify our research by use of telephone questionnaire for residents in Kunsan, Republic
of Korea.
․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․․
Received : December 18, 2013
Accepted : December 24, 2013
Type of Submission : Concise Paper Corresponding author : Sangwon Lee
J Intell Inform Syst 2013 December: 19(4): 11~20
ISSN 2288-4866 (Print)
ISSN 2288-4882 (Online)
http://www.jiisonline.org
http://dx.doi.org/10.13088/jiis.2013.19.4.011
* This
p
a
p
er was su
pp
orted
by
Wonkwan
g
Universit
y
in 2013.
1. Introduction
Intelligent Traffic Systems involve an effi-
cient and effective interaction between all of its
components such as drivers, pedestrians, traffic
management systems, and public transportation.
Intelligent traffic systems provide various in-
telligent services (e.g. Big Data Analysis or
Data Mining) (Baek et al., 2010; Kim et al.,
2010; Cho et al., 2011; Yu et al., 2013) by aim-
ing to make full use of car properties. Especially
in car safety, the car color is a very important
factor and should be considered in designing
and implementing intelligence services as well
Seong-yoon Shin
․
Sangwon Lee
12
지능정보연구
제
19
권 제
4
호
2013
년
12
월
as in purchasing a car.
Let us assume that a customer considers
purchasing a car. There are five major factors in
selecting a car. We first of all consider its use
for private, family, and freight. There is no ne-
cessity for buying a big or expensive car for in-
dividual commuting. Secondly, we consider its
price with guarantee condition, installment plan,
and so forth. Thirdly, we analyze the price and
performance of the selected car. The analysis
could be supported by preference reports or ex-
pert opinions on that car. Fourthly, we try to
drive and operate the car. This try is a good fac-
tor in experiencing the car. Fifthly, after these
four considerations, we check the price again. In
the ear of high oil price, we never even dream
of driving a car. So, we choose a car after
checking its average mileage a liter. And we al-
so consider whether we will drive it in the town
or on the high ways as well as on the paved
road or on the unpaved road. Lastly, we consid-
er its design and color that we like. However,
we should know that the selected color is close-
ly related with car accident.
Since car color has something to do with
car accident, the relationship between accident
and color is a very important factor. There are
hundreds of car body colors as car flame.
However, only eight colors are widely used. We
limit 8 car colors in order to express relationship
between car accident and car color. In Korea,
deep colors like black were widely used in the
days when cars were rare. In these days of
Korea, light colors like silver, gold, and white
are largely used. Lately, cars in white, gold (or
yellow), and silver are increasing more and
more. What color of cars was used the most 2~3
years ago in Korea? It is not difficult to know
the colors. In Korea, black, white, and silver are
most widely used. It would be so since many
Koreans like quite tones. But, from the last
2011, colors of cars have been changing little by
little. Small-sized cars show this phenomenon
remarkably rather than large-sized ones. Golden
or yellowish colors as well as pinkish ones have
been increasing. In fact, sales volume of yellow-
ish and pinkish cars actually increased. Even
though this phenomenon shows the expressions
of personality and tendency of car owners, it is
considered to have anything to do with car
accident.
This paper studies on the rate or ranking
according to car colors in Kunsan, Republic of
Korea. Session 2 shows related works on rela-
tionship between car color and car accident and
Session 3 checks concepts of advancing color
and receding color. In Session 4, we perform an
experiment for relationship between car color
and car accident. Lastly we make conclusions
and further studies in Session 5.
2. Related Works
The damage extent of car accident and so-
cial cost are as follows. According to data of
Korea Road Traffic Authority in <Table 1>,
Correlation between Car Accident and Car Color for Intelligent Service
지능정보연구
제
19
권 제
4
호
2013
년
12
월
13
there are 897,271 car accidents in 2011 with
killing 5,229 and wounding 1,434,786 (seriously
wounded 173,809, slightly wounded 935,449,
and wounded reported 325,528.
<Table 1> Year-on-Year Traffic Accident Overview
(Unit: Person, %)
Basic Year 2011 2010
Number of Cases 897,271 979,307
Increasing/Decreasing Number -82,036 -
Increasing/Decreasing Rate -8.4 -
Killed 5,229 5,505
Increasing/Decreasing Number -276 -
Increasing/Decreasing Rate -5 -
Injured 1,434,786 1,533,609
Increasing/Decreasing Number -98.823 -
Increasing/Decreasing Rate -6.4 -
Source : KoROAD.
The yearly cost of road accidents in Korea
increases to 13 trillion won. KoRoad said the
social cost that is caused by car accident of
2010. The cost of 2010 is 12,959.9 billion won,
which is 1.1 percent of GDP or 6.4 percent of
a national budget of Korea. The cost of 2009 in-
creases 10 percent (118.24 billion won) more
than that of 2009. This scale is 1.1 percent of
GDP 1172.8034 trillion won or 6.4 percent of
Korea national budget 201.2835 trillion won.
The Shin et al. (Shin et al., 2013) shows the re-
lationship between car color and car accident on
the basis of chromatic aberration. But, we per-
form an experiment for 138 cars by use of seven
colors. It is not easy to show the relationship
since the number of data is very small. The Fur-
ness et al. (Furness et al., 2003) investigated the
effect of car color on the risk of a serious injury
from a crash, using a population based case con-
trol study designed to identify and quantify mo-
difiable risk factors. According to Ansah et al.
(Ansah et al., 2010), over the years, the concern
of many, consumers and insurance companies
alike, has been geared towards the contribution
of vehicle color to the risk of crash. Conse-
quently, there is a need to provide sufficient sci-
entific evidence to back consumers in selecting
the appropriate vehicle color that enhances their
safety on the road. Very little research has been
conducted to study whether vehicle color may
have an effect on motor vehicle crash. Particu-
larly, scientific studies to determine the relation-
ship between vehicle color and crash risks have
been scarcely investigated (Newstead, et al.,
2007). Many studies have investigated the rela-
tionship between color and visibility (FEMA, et
al., 2009) and most of them have focused on re-
flectivity of sign visibility (Anders et al., 2000;
Hawkins et al., 2000; Gates et al., 2004).
3. Concepts of Car Color
The retraction index of color and focusing
function of our eyes make location of an object
look differently according to color. In chro-
matics, this phenomenon is called as Chromatic
Aberration. Chromatics is a science that determi-
nes the essence of chromatic phenomena and re-
searches on mutual relationships between color
and human living. Let us assume that a black
car and a golden car stand in the same location.
Seong-yoon Shin
․
Sangwon Lee
14
지능정보연구
제
19
권 제
4
호
2013
년
12
월
The black car looks smaller as if it would be far
from its real location. Meanwhile, the golden car
looks closer than it really is. Since the retraction
index of light is small in case of gold chromatic
aberration, the image is focused behind the
retina. Then, the eyeball inflates crystalline lens
in order to focus the image on the retina. At this
moment, the crystalline lens inflates and becomes
convex lens. So, the golden car looks closer
than it really is. On the contrary, the black car
looks farther than it really is. In addition, there
are advanced color and receding color in the
world. An advanced color means the color that
looks closer although it really is farther. A re-
ceding color means the color that looks farther
although it really is closer. In this paper, we as-
sume that the concept of chromatic aberration is
the same as that of advanced color and receding
color.
3.1 Advancing Color
An advanced color means the color that
looks closer although it really is farther. Warm
colors such as red or yellow belong to the ad-
vanced color. In other words, advanced color is
the color that looks about to stick out forward
rather than background color. Colors with high
brightness and high chroma look closer although
they really are farther. We can easily find out
the effect if colors are arranged by stages.
Advanced color is also called expansive image.
Advanced colors are yellow (gold), brown
(chestnut), black, and so on. In <Figure 1>, they
look larger than other colors with expansive pro-
perty.
<Figure 1> Advanced Colors (Order : Left-Top →
Right-Top, Left-Bottom → Right-Bottom)
3.2 Re c e di ng Co l o r
A receding color means the color that
looks farther although it really is closer. Colors
have low brightness and low chroma. We can
easily find out the effect if colors are arranged
by stages. Receding color is also called con-
tractive image. Receding colors are green, blue,
and so on. Adversely, receding color is the color
that looks about to draw back backward in
<Figure 2>.
<Figure 2> Receding Colors (Order : Left-Top →
Right-Top, Left-Bottom → Right-Bottom)
Warm colors such as yellow and red are
advanced colors since they look about to stick
out forward rather than cold colors such as blue
Correlation between Car Accident and Car Color for Intelligent Service
지능정보연구
제
19
권 제
4
호
2013
년
12
월
15
and blue-green. Cold colors are receding colors
since they look farther. Chromatic colors tend to
advance rather than achromatic ones. In case of
dark backgrounds, lighter colors are more ad-
vanced. But, in case of bright backgrounds, dar-
ker colors are more advanced. Warm colors are
more advanced than cold ones. Bright colors are
more advanced than dark ones. Colors with high
chroma are more advanced than those with low
chroma. Although white, red, black, and silver
are cold or warm colors in <Figure 1> and
<Figure 2>, they tend to look different from the
standard of chromatic or achromatic color. But,
from the enamel color characteristic of real cars,
the order of figure is right. Enamel color is
quicker drying than oil paint. And its Dry-Paint-
Film is smoother and its gloss is more fluent.
4. Experiment : Rate of Accident and
Ranking according to Car Color
In choosing a car, most of people prefer-
entially consider the usage and price of the car
among various conditions such as option, install-
ment condition, and so on. But, we should know
that color is an important factor to be consi-
dered. Car color has something to do with car
accident and is a very important factor. For safe-
ty, it is better to choose expansive color with
bright tone than to choose contractive with dark
tone. So as to check relationship between car
color and car accident, we should choose wide-
ly-used colors. <Figure 3> shows the arrange-
ment of representative car colors. The colors are
classified into eight colors from blue to yellow
(gold). Seven colors of them go match car colors.
These eight colors represent a lot of colors. For
instance, the colors with red tone are commonly
called ‘red.’ The detailed colors are named by
car manufactures so we leave the names out of
discussion.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
<Figure 3> Eight Car Colors
Now, we show four practical examples with
the above-mentioned three assumptions, five
definitions, and nine rules with four ones to ex-
tract objects and associations and five ones to
extract entities and relationships.
We performed an experiment with crapped
10-old-year cars in Kunsan. We checked scrap-
ped cars in various locations such as junkyards
of local the government for research on colors
of car accidents. But there were no materials re-
lated to our research concerns. So, we should
check colors of scrapped cars or perform phone
interviews with individuals for ourselves. Table
2 shows the questionnaire performed in phone
Seong-yoon Shin
․
Sangwon Lee
16
지능정보연구
제
19
권 제
4
호
2013
년
12
월
or direct interviews.
<Table 2> Questionnaire on Car Accident
Items of Questiionnaire Remarks
Is there an accidentor not? Yes or No
Which is the accident car,
sedan or van? Sedan of Van
How many accidents do you
have? One to ten
What color of the accident
car is?
Blue, Green, White, Red,
Black, Silver, Brown, and
Yellow (Gold)
Is there any relation
between the accident car
and its color?
Yes or No
500 persons responded to the 1123 questi-
onnaires. We checked accident rates and ranking
for each color by use of 500 responses. We got
a result like <Table 3>.
<Table 3> The Number of Car Accidents
Accident Yes No
The number of cars 500 623
Percentage 44.5 55.5
Our questionnaire includes an item in or-
der to check the type of accident cars, sedan or
van <Table 4>. The sedan is used to transport
persons, and the van is used to transport fre-
ights.
<Table 4> The Type of Car Accidents
Type Sedan Van
The number of cars 324 176
Percentage 64.8 35.2
There are cars with at most 10 accidents.
Some cars have no accident. We calculate the
number of accidents redundantly for each car. In
addition, we researched on the color of all acci-
dent cars. <Table 5> shows the preference of
car color. Silver has the highest preference and
brown (chestnut) is the lowest one.
<Table 5> The Preference of Car Colors
Car Color The Number
of Vehicle
Preference
of Color
Preference
Rank
Blue 42 8.4 4
Green 36 7.2 6
White 104 20.8 2
Red 38 7.6 5
Black 96 19.2 3
Silver 114 22.8 1
Brown (Chestnut) 34 6.8 8
Yellow (Gold) 36 7.2 6
Total 500 100 -
<Table 6> shows the number of car acci-
dents according to car colors. The ratio of acci-
dent cars is calculated with the number of car
accidents divided by the number of cars. The
more the ratio is, the more the accidents are. In
<Table 6>, the rank of blue is 1 since it is a re-
ceding color. The rank of yellow (gold) is 8 so
its probability of car accident is the least. Since
yellow has no chromatic aberration, it is focused
on the retina when entering eyes. Yellow has
properties to expand on the retina like ink and
look larger among colors. So, for drivers, it is
used as hats, raincoats, backpacks of preschool
and elementary school. The descending order of
accident occurrence probabilities is blue, green,
Correlation between Car Accident and Car Color for Intelligent Service
지능정보연구
제
19
권 제
4
호
2013
년
12
월
17
white, red, black, silver, and yellow (gold).
<Table 6> Ratio of Car per Accident and Rank
Car Color The Number
of Cars
The Number
of Accidents
Ratio of
Accidents
per Car
Rank
Blue 42 145 345.24 8
Green 36 119 330.56 7
White 104 301 289.42 6
Red 38 105 276.32 5
Black 96 259 269.79 4
Silver 114 274 240.35 3
Brown (Chestnut) 34 80 235.29 2
Yellow (Gold) 36 76 211.11 1
Total 500 1,359 - -
In <Table 7>, 309 responders (79 percent)
think that there is relationship between car color
and car accident. The number of the responders
is four times as many as that of rest (21 percent)
who does not think any relationship between car
color and car accident. Consequently, most of
drivers think that color has something to do with
accident.
<Table 7> Relationship between Car Color and Car
Accident
Relationship Sedan Van
Persons 395 105
Percentage 79 21
5. Conclusions
For traffic safety in making car-purchasing
decision or administrating Intelligence Traffic
Systems, color should be considered on the pref-
erential basis before other options. As we al-
ready performed telephone questionnaire for citi-
zens in Kunsan in order to research on car acci-
dents and their car colors, the results are as
follows. Advanced colors have low probability
of car accident since they look closer than they
really are. On the contrary, receding colors have
high probability of car accident since they look
farther than they really are. And we classified
car colors into eight ones such as black, white,
blue, green, silver, red, brown (chestnut), and
yellow (gold). After considering the probability
of car accidents, we concluded that the descend-
ing order is blue, green, white, red, black, silver,
brown (chestnut), and yellow (gold). And we
proposed the usage of cars, the ratio of sedan
versus van, the preference of car color, and the
ratio of car color and car accident. However, al-
though we found out the fact that car color has
something to do with car accident in the field of
safety, more consideration of other properties
should be researched further.
References
Anders, R. L., “On-road Investigation of Fluore-
scent Sign Colors to Improve Conspicuity,”
Virginia Polytechnic Institute and State Uni-
versity, 2000.
Ansah, O. and S. Osei, “Investigation of the Rela-
tionship between Vehicle Color and Safety,”
Master of Science, University of Dayton,
Department of Civil Engineering, 2010.
Baek, W. and N. Kim, “A Case Study on Fore-
casting Inbound Calls of Motor Insurance
Seong-yoon Shin
․
Sangwon Lee
18
지능정보연구
제
19
권 제
4
호
2013
년
12
월
Company Using Interactive Data Mining
Technique,” Journal of Intelligence and
Information Systems, Vol.16, No.3(2010),
99
~
120.
Cho, I. and N. Kim, “Recommending Core and
Connecting Keywords of Research Area
Using Social Network and Data Mining Te-
chniques,” Journal of Intelligence and Infor-
mation Systems, Vol.17, No.1(2011), 127~
138.
FEMA(Federal Emergency Management Agency),
Emergency Vehicle Visibility and Conspicuity
Study, U.S. Department of Homeland Secu-
rity, 2009.
Furness, S., J. Connor, E. Robinson, R. Norton,
S. Ameratunga and R. Jackson, “Car Colour
and Risk of Car Crash Injury: Population
Based Case Control Study,” British Medical
Journal, Vol.327, No.7429(2003), 1455~
1456.
Gates, T. J. and H. G. Hawkins, “Effect of Hi-
gher-Conspicuity Warning and Regulatory
Signs on Driver Behavior,” Texas Transpor-
tation Institute Report, No. 0-4271-S(2004),
Texas A&M University.
Hawkins, H. G., P. J. Carlson and M. Elmquist,
“Evaluation of Fluorescent Orange Signs,”
Texas Transportation Institute Report, No.
0-2962-S(2000), Texas A&M University.
Kim, M., N. Kim and J. Ahn, “An Investigation
on Expanding Co-occurrence Criteria in As-
sociation Rule Mining,” Journal of Intelli-
gence and Information Systems, Vol.18, No.1
(2010), 23~38.
Newstead, S. and A. D’Elia, “An Investigation
into the Relationship between Vehicle Color
and Crash Risk,” Accident Research Centre
Report, Monash University, Vol.263(2007).
Shin, S., Y. Rhee, D. Jang, S. Lee, H. Lee and
C. Jin, “Relationship Between Car Color and
Car Accident on the Basis of Chromatic
Aberration’, Future Information Communi-
cation Technology and Applications,” Lec-
ture Notes in Electrical Engineering, Vol.
235, No.1(2013), 45~51.
Yu, E., J. Kim, C. Lee, N. Kim and S. Jeong,
“Predicting the Direction of the Stock Index
by Using a Domain-Specific Sentiment Dic-
tionary,” Journal of Intelligence and Infor-
mation Systems, Vol.19, No.1(2013), 95~110.
Correlation between Car Accident and Car Color for Intelligent Service
지능정보연구
제
19
권 제
4
호
2013
년
12
월
19
Abstract
지능형서비스를 위한 자동차사고와 자동차색깔의 상관관계
1)신성윤*․이상원**
지능형 교통 시스템을 설계하는데 있어서
,
통신
,
외관
,
환경
,
부수 기능
,
안전 등이 반드시 고려되어야 한다
.
또한
,
자동차를 선택하는데 있어서도
,
소비자는 이러한 특성들을 고려한다
.
본 논문에서는
,
지능화된 교통 서비스를 운영하거나
자동차 구매 의사결정을 하는데 있어서
,
자동차 색깔이 자동차 안전에 있어서 매우 중요한 의미를 갖는다는 것을 밝히고자
한다
.
자동차 안전과 관련이 있는 자동차 색깔에 따른 자동차 사고 확률에 대해 연구하였다
.
그리고
,
전진색과 후퇴색의
개념에 대해서도 연구하였다
.
전진색은 실제보다 가깝게 보이기 때문에 사고를 적게 유발하지만
,
후퇴색은 실제보다 멀리
보이기 때문에 사고를 많이 유발한다
.
우리는 자동차 사고 건수를 고려하여
,
자동차 색깔을
8
개로 분류하고 이들의 등급을
부여하였다
.
또한
,
우리의 연구를 실증하기 위해서
,
군산에 있는 거주자를 대상으로 전화상담을 통해 조사를 하였다
.
Keywords : 지능형교통시스템, 자동차색, 자동차사고, 전진색, 후퇴색, 색수차
*
군산대학교 컴퓨터정보공학과
**
원광대학교 정보전자상거래학부
Seong-yoon Shin
․
Sangwon Lee
20
지능정보연구
제
19
권 제
4
호
2013
년
12
월
저 자 소 개
Seong-yoon Shin
Seong-yoon Shin (the first author) received his Ph.D. degree in Computer Science from
Kunsan National University in 2003. Since 2006, he has been working as an associate
professor for Kunsan National University. His research interest is computer engineering
such as computer vision, image processing, and so on.
Sangwon Lee
Sangwon Lee (the corresponding author) received his Ph.D. degree in Management
Engineering from Korea Advanced Institute of Science and Technology in 2009. Since
2011, he has been working as an assistant professor for Wonkwang University. His research
interest is management engineering such as data engineering, data science, and so on.