In e na ional Jou nal o Applied Enginee ing Resea ch ISSN 0973-4562 Volume 11, Numbe 9 (2016) pp 6373-6375
© Resea ch India Publica ions. h p://www. ipublica ion.com
6373
De e mina ion o he Unce ain ies and he Physiological Simila i ies o
Family Membe s by using he Biome ic De ice he B oadway 3D
Hana Talando a, Lukas K alik, Milan Adamek
Secu i y Enginee ing Depa men ,
Tomas Ba a Uni e si y in Zlin, Facul y o Applied In o ma ics, Zlin, Czech Republic
Abs ac
The biome ic iden i ica ion by he ace is one o he oldes
biome ic iden i ica ion. Wi h inc easing p og ess has been
using o iden i ica ion by he ace was implemen ed in o a ea
o secu i y, whe e i p o ides a as e and mo e accu a e
iden i ica ion. The 3D ace eade uses o he iden i ica ion o
he pe son: eyes, mou h, nose, and in con as o 2D eade s
also chin and cheeks. 3D ace eade by B oadway
manu ac u e was used o measu e he physiological
simila i ies o amily membe s. I is equipped wi h he 3D
came a sys em, which uses he me hod o s uc u ed ligh
scanning and sa es he empla e in o he 3D model o ace.
The ob ained da a we e e alua ed by so wa e Tu ns ile
En olmen Applica ion (TEA). The pa icipan s o he
measu emen we e membe s o h ee di e en amilies. Each
pe son was compa ed wi h he p e iously sa ed empla es o
o he amily membe s. On basis o his ac was e alua ed he
simila i y o amily membe s.
Keywo ds: Biome ic iden i ica ion, ace, B oadway 3D
eade , simila i y
In oduc ion
The iden i ica ion o people by ace is one o he oldes
me hods o biome ic iden i ica ion. People ecognizing
ano he pe son by ace, because i is he easies and he mos
na u al way. The ace is sui able o iden i ica ion o
e i ica ion. This me hod was g adually in oduced in o he
secu i y a ea. A he en ance o he objec , he use submi s
iden i ica ion documen o he doo man who subsequen ly
ecognized by he a ached pho o o ace, whe he i is he
au ho ized pe son. This ype o e i ica ion is s ill used. . Wi h
he p og ess o echnology, doo men ha e been eplaced by
his biome ic sys em. The secu i y s a can make a mis ake
when hey check a la ge numbe o people. The biome ic
eade o ace p o ides a as e and mo e accu a e
iden i ica ion hen s anda d isual me hod.
The ace is composed om h ee basic cha ac e is ics he
ex e nal s uc u es, which help o shape he o e all pic u e o
he ace. These a e he eyes, he nose and he mou h. These
pa s a e a iables o each e hnic g oup, i is educes he
likelihood o wo people wi h all iden ical cha ac e is ics.
O he cha ac e s de e mining he ace image a e chin and
cheeks, which a e impo an especial o 3D ace eade s. The
shape o he chin is loca ed in he lowe pa o he ace and i
is de e mines he o e all shape o he lowe pa o he ace.
The cheek is ano he pa which is impo an especially o
side iew. I is o med by he shape o he ob e se, he
empo al bone and hei in e connec ion cheeks a ches.
Fu he mo e, i s shape is a ec ed by so pa s o he ace,
hus acial muscles and subcu aneous issue.
Facial Recogni ion by Using he 3D Reade s
The ace is composed om h ee basic cha ac e is ics he
ex e nal s uc u es, which help o shape he o e all pic u e o
he ace. These a e he eyes, he nose and he mou h. These
pa s a e a iables o each e hnic g oup, i is educes he
likelihood o wo people wi h all iden ical cha ac e is ics.
O he cha ac e s de e mining he ace image a e chin and
cheeks, which a e impo an especial o 3D ace eade s. The
biome ic sys ems o iden i ying people by 3D images a e
sa e and mo e e ec i e me hod o iden i ying. 3D ace
image is c ea ed by using he polygonal ne wo k o iangula
shape.
Figu e 1: The p ocess o acial ecogni ion
Face De ec ion consis s o ou s eps (Figu e 1). The i s s ep
is o iden i y he ace in he image. The eyes and mou h a e
sea ched in images ha oge he hey o m a iangle and hey
a e da ke han hei su oundings. The p ocess con inues by
no maliza ion ace, so ha he image o ace was p ede ined
s anda dized dimensions and i was compa able wi h a pa e n.
Whe e he acial image is c opped and s anda dized in e ms
o size, pose, ligh ing o he s o ed empla e ace. Then hey
a e ound indi idual cha ac e is ics ha a e obse ed ( egions
eyes, nose, mou h, chin, eyeb ows and o he impo an poin s
on he ace). A he end o he ou h s ep is he ecogni ion o
aces. [1,2]
The Biome ic De ice B oadway 3d
The biome ic de ice B oadway 3D ep esen s he possibili y
o p o ec ion objec s which equi e a high secu i y le el. The
de ice acqui es biome ic samples ( empla e) o he use a
egis a ion, which a e used o iden i ica ion o e i ica ion.
The B oadway in comple e au onomy e alua es whe he he
use is au ho ized o no . The esul ing in o ma ion is
displayed on he connec ed display uni o indica ed by LEDs
in he uppe pa o he de ice[7].
In e na ional Jou nal o Applied Enginee ing Resea ch ISSN 0973-4562 Volume 11, Numbe 9 (2016) pp 6373-6375
© Resea ch India Publica ions. h p://www. ipublica ion.com
6374
B oadway 3D ace eade is i s de ice a ound he wo ld
which is capable o isually iden i y pe son as easily as people
a e iden i ying wi h each o he . While he iden i ica ion only
akes a spli second. Sigh on he de ice du ing walking o
e en unning is su icien o enable The B oadway o
ecognize he iden i y o one use o he ens o housands o
p e iously egis e ed indi iduals. The 3D ace eade The
B oadway is equipped by 3D came a sys em which eco ds
and sa es he unique h ee-dimensional shape o ace. In
con as o he human eye he de ice has he abili y o
dis inguish sligh he di e ences o geome y wi h accu acy o
ac ion o a millime e, allowing i o de ec e en iden ical
wins. Technology o 3D scanning enables as and p ecise
scanning he aces o pe son by using he me hod o s uc u ed
ligh . The ligh is p ojec ed h ough he g id on he su ace o
he objec whe e he came a scans he su ace o he objec
om ano he loca ion han he ligh and hus leads o
measu ing he angle and he dis ance. The ad an age o 3D
ace de ec ion consis s in he abili y o measu e he angle and
he dis ance, unlike he 2D eo ganiza ion which only measu e
he dis ance in he s uc u e o ace. The g id o med on he
su ace o he scanned objec he ne wo k coo dina es, which
a e used o accu a e calcula ion o each poin . The ex u e o
came a is synch onized wi h he 3D senso , which has a wide
ield o iew. The senso cap u es he shape and su ace
s uc u e o he objec . F ame equency o he came a is 15
ames pe second and cap u es images a wide angles up o
41 deg ee. So i is able o cap u e people's aces eaching
down hal he heigh o an adul such as child en and people
in wheelchai s [7].
Me hodology o Measu emen
Fo wo king wi h he biome ic de ice was used eewa e
so wa e Tu ns ile En olmen Applica ion (TEA). The
applica ion TEA was de eloped o managing o biome ic
da abase. The TEA is use - iendly and o e s many
possibili ies. I ep esen s easie al e na i e o SDK, which
also allows he biome ic da a om he eade .
The B oadway was connec ed o he wo compu e s. The
compu e 1, which includes a da abase o biome ic samples,
is connec ed wi h he compu e 2 by using E he ne cable. The
B oadway 3D is connec ed di ec ly o he compu e uni 2 by
using USB. The compu e 1 keeps he da abase o biome ic
samples and communica es wi h a TEA, which is ins alled on
he compu e 2[6].
Figu e 2: Comple e Measu emen In e ace-Schema ic[4]
The h eshold was speci ied 60%, hus he lowes possible
alue, which causes a eac ion o he biome ic sys em. When
applies ha he smalle h eshold, he g ea e p obabili y o
he alse iden i ica ion o e i ica ion o he pe son. And wi h
high h eshold he pa icula use may no be iden i ied. In his
case, depends on he quali y o he model, gene a ed du ing
scanning aces which is a ec ed by en i onmen .
Du ing egis a ion he use is asked o appea be o e he
B oadway 3D and main ain a speci ied dis ance om he
biome ic de ice om 90 o 120 cen ime es. The quali y o
he model c ea ed du ing he egis a ion p ocess can be
e alua ed based on he alue o quali y he egis a ion. This
pa ame e can ake alues be ween 0 and 1000. The model
was e alua ed as sui able o use as a e e ence empla e,
when he alue anged om 850 o 1000. O he wise, he
egis a ion had o be done again. Subsequen ly, he
iden i ica ion was pe o med o each pe son and compa ing
wi h he each amily membe .
Resul o measu emen he physiological simila i ies o amily
membe s
The Resul o Measu emen
The measu emen was pa icipa ed by h ee amilies. The i s
amily is h ee gene a ions and has 8 membe s. While he
second and hi d amily is only wo gene a ions which ha e
only 3 and 4 amily membe s. Fi s , he biome ic empla e
was c ea ed o all measu ed pe sons. The ea e , each pe son
was measu ed and compa ed wi h hei empla e (yellow
alues) and subsequen ly wi h o he amily membe s.
Table 1: Family 1
GMS
GMI
Fa he
Mo he
Aun
Son1
Son2
Cousin
GMS
60.4
22.2
32.6
18.7
19
G MI
60.3
15.8
14.7
14
16
Fa he
18.4
61.2
20.3
25.3
21.2
Mo he
12.2
15.1
14.1
69.3
13.4
Aun
24.1
60.4
21.6
18.6
24.9
Son1
16.5
21.3
19.9
63.2
23.1
Son2
16.2
20.1
16.6
22.3
69.3
Cousin
18.5
16.5
14.7
22.3
60.2
Table 1 shows he amily 1. The amily lineage in his
in es iga ed sample is G andmo he S (GMS) has a son and a
daugh e ( a he and aun ) G andmo he I (GMI) has a
daugh e (mo he ), a he and mo he ha e 1 son, bu his
a he s ill has a son 2.
Compa ing o sa ed empla es be ween a he -son b ings
simila i y in he ange om 20.1% o 25.3%. Compa ing o
sa ed empla es be ween mo he -daugh e b ings simila i y in
he ange om 14.7% o 32.6%. The deg ee o simila i y
be ween siblings is a ound 23%, whe eas compa ing wi h he
o he amily membe s, he deg ee o he simila i y is a ound
15%.
In e na ional Jou nal o Applied Enginee ing Resea ch ISSN 0973-4562 Volume 11, Numbe 9 (2016) pp 6373-6375
© Resea ch India Publica ions. h p://www. ipublica ion.com
6375
Table 2: Family 2
Son
Fa he
Mo he
Son
70
Fa he
23
83.7
Mo he
70.8
In Table 2, we can see a amily 2 whe e alue du ing epe ed
iden i ica ion o he same pe son was ob ained in he ange
om 70% o 83.7% .Acco ding o ob ained alues; he highes
simila i y is be ween a he and son. When, hei ela i e
simila i y is in he ange om 23% o 25.4%.
Table 3: Family 3
Mo he
Fa he
Son1
Son2
Mo he
62.1
Fa he
61.1
27.8
Son1
28.8
60.9
29.7
Son2
27.3
23.9
78.3
Table 3 p esen s alues o a wo-gene a ion amily 3, which
consis s o ou membe s. He e we can see again he ela i e
simila i y be ween sons and a he . When, he alue o
simila i y is in he ange om 23.9% o 29.7%. The deg ee o
simila i y be ween siblings is in he ange om 23.9% o
29.7%.
Tables show he compa ison o he physiological simila i ies
o amily membe s in he biome ic iden i ica ion by he ace.
As we can see, epea ed iden i ica ion o he same pe son is in
he ange om 60% o 83% compliance wi h he sa ed
empla e. This low compliance may be caused by ad e se
ligh condi ions, when measu emen was made in illumina ion
wi h luo escen lamps. Compa ing o sa ed empla es
be ween a he -son b ings simila i y in he ange om 20% o
30%. Compa ing o sa ed empla es be ween mo he -daugh e
b ings simila i y in he ange om 14% o 32%. The deg ee o
simila i y be ween siblings is also a ound 25%, whe eas
compa ing wi h he o he amily membe s, he deg ee o he
simila i y is a ound 15%. In Tables, he e a e some emp y
cells because simila i y was lowe han 10% and his alue is
i ele an . Regis a ion was signi ican ly mo e di icul o
women who ha e long hai . In many cases, he hai all down
in o his ace and educe he isible a ea aces. This is e lec ed
in he quali y o he esul ing empla e. The e o e, i was o en
necessa y o epea he egis a ion p ocess.
Conclusion
The measu ed esul s show ha he deg ee o simila i y in he
amily is maximal 32%. When, he simila i y in he amily
he e is always he mo he -daugh e and a he -son. Fu he ,
he measu ed alues show ha i amily membe s o e wo
gene a ions a e compa ed, hen he deg ee o simila i y is low.
Howe e , when wo amily membe s o he same gene a ion
o ollowing gene a ion a e compa ed, hen he esul s o
simila i y a e highe .
The biome ic accesses a e inc easingly also used in he public
a eas and he e o e p ecision and accu acy o scanning is e y
impo an . The e o e, i is necessa y o se up co ec ly he
h eshold in he access sys ems in o de o a oid g an ing
access o an unau ho ized pe son.
Acknowledgmen
Wi h suppo by g an No. IGA/FAI/2016/027 and
IGA/CebiaTech/2016/006 om IGA (In e nal G an Agency)
o Thomas Ba a Uni e si y in Zlin. This wo k was suppo ed
by he Minis y o Educa ion, You h and Spo s o he Czech
Republic wi hin he Na ional Sus ainabili y P og amme
p ojec No. LO1303 (MSMT-7778/2014).
Re e ences
[1] W. Zhao, R. Chellappa, A. Rosen eld, and P. Phillips.
Face ecogni ion: A li e a u e su ey. ACM Compu ing
Su eys, pp. 399-458, 2003.
[2] Ma qu´es. Face ecogni ion algo i hms. P oyec o Fin de
Ca e a, pp. 1-78, 2010
[3] Rak, Roman, Václa MATYÁŠ a Zdeněk ŘÍHA.
Biome ics and iden i y o a pe son in o ensic science
and comme cial applica ions. 1. P aha: G ada
Publishing, a.s., 2008. ISBN 978-80-247-2365-5.
[4] LI, Haizhou, Liyuan LI a Ka -Ann TOH. Ad anced
opics in biome ics. New Je sey: Wo ld Scien i ic,
c2012, x , 500 s. ISBN 978-981-4287-84-5.
[5] Biome ics [online]. 2007 [ci . 2010-05-16]. Dos upnéz
www :<h p://pagespe so-o ange. / inge chip/bio
me ics/biome ics.h m>.
[6] A ec G oup, Inc. Tu ns ile En ollmen Applica ion:
Use Manual. Luxembou g, 2011.
[7] ARTEC GROUP, Inc. B oadway 3D: Use Manual.
B no, [2011].
[8] Talando a, Hana. S udy abou applica ion o biome ic
sys ems in he indus y o comme cial secu i y. Zlin,
2010. Bachelo hesis. UTB in Zlin.
[9] D ahanský, Ma in a Filip ORSÁG. Biome ics. 1.
[B no: M. D ahanský], 2011, 294 s. ISBN 978-80-254-
8979-6.
[10]