Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
V
o
l.
6, N
o
. 3
,
Ju
n
e
201
6, p
p
. 1
152
~ 11
60
I
S
SN
: 208
8-8
7
0
8
,
D
O
I
:
10.115
91
/ij
ece.v6
i
3.9
233
1
152
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Stone I
m
age Cl
assifi
cati
on
Based on Overlapped 5-bit
T-Patterns Occurrence on 5-by-5 Sub Images
Pa
lna
t
i
Vija
y
Kuma
r, Pullela
S V V S
R
Kumar
,
Nakkel
la Madhuri,
M
Um
a Devi
Department o
f
C
o
mputer Scien
c
e &
Engi
ne
ering
,
Adit
ya Co
lleg
e
of Engin
eerin
g,
S
u
ram
p
alem
, An
dhra P
r
ades
h
,
In
dia
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Sep 20, 2015
Rev
i
sed
No
v 6, 201
5
Accepted Nov 26, 2015
Textur
e classification is widely
u
s
ed in
understan
d
ing the visual p
a
tterns and
has wide range
of applications
. The
present p
a
p
e
r derived
a novel approach
to clas
s
i
f
y
th
e s
t
one tex
t
ures
ba
s
e
d on the pat
t
e
r
ns
occurrenc
e
on each s
u
b
window. The pr
esent appro
ach
identif
ies over
l
apped nine 5 b
i
t T-patterns
(O5TP) on each 5×5 sub window stone
image. Based
the
number of
occurren
ce of
T-patterns count
the pres
ent paper classif
y
the stone images
into an
y
of th
e
four classes i.e. bric
k
,
granite, marble and mosaic stone
im
ages
.
The no
velt
y of
th
e pres
ent app
r
oach
is
t
h
at no s
t
andard
c
l
as
s
i
fic
a
tio
n
algorithm is used for the
classification of stone ima
ges. The prop
osed method
is experimented
on May
a
ng
tex
t
ure imag
es, Br
odatz textur
es,
Paul Bourke
color images, V
i
sTex database,
Google
color stone textur
e images and also
original pho
to images taken b
y
digital
camera. The outcome of the results
indicates th
at
th
e proposed appr
oach percen
tag
e
of grouping per
f
ormance is
higher to
th
at
of man
y
existing
approaches.
Keyword:
Grey lev
e
l im
a
g
e
Im
age classification
Stone
im
ages
Texture Classification
T-p
a
ttern
s
Copyright ©
201
6 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Dr.
P
u
llela S
V
V
S R
Kum
a
r,
Depa
rt
em
ent
of C
o
m
put
er
Sci
e
nce &
E
ngi
ne
eri
n
g,
Adi
t
y
a C
o
l
l
e
ge
o
f
E
ngi
neeri
n
g,
Suram
p
ale
m
, Andhra
Pra
d
es
h,
India.
Em
a
il: p
u
llelark
@yah
oo
.co
m
1.
INTRODUCTION
TEXT
U
R
E
an
al
y
s
i
s
and cat
ego
r
i
zat
i
on a
r
e
im
port
a
nt
f
o
r
t
h
e i
n
t
e
rp
ret
a
t
i
on an
d u
n
d
er
st
andi
ng
o
f
real-world vis
u
al patterns. T
e
xture cl
assi
fi
cat
i
on has a wi
de va
ri
et
y
of p
r
os
pect
i
v
e a
p
p
l
i
cat
i
ons [1]
s
u
ch as
reg
i
o
n
s
classificatio
n
in satellite i
m
ag
es [2
],
d
e
fects
d
e
tectio
n
in indu
strial su
rface in
sp
ectio
n [3], and
cl
assi
fi
cat
i
on o
f
pul
m
onary
di
sease
[
4
]
,
di
ag
nosi
s
o
f
le
uke
mic cells in medical im
age [5] and
breast c
a
ncer
classification [6]. Text
ure a
n
alysis and cl
assificati
on i
s
m
a
jorl
y
achi
e
ved i
n
o
n
e
o
f
t
h
e t
w
o wa
y
s
, i
.
e.
st
at
i
s
t
i
cal appr
oach a
n
d st
r
u
c
t
ural
m
e
t
hod.
St
at
i
s
t
i
cal
appr
oach m
a
i
n
l
y
conce
n
t
r
at
es
on
t
h
e st
och
a
st
i
c
t
h
i
n
g
s
o
f
th
e sp
atial
distrib
u
tion
o
f
gray lev
e
ls i
n
an
im
ag
e.
Ge
ne
rally
fo
r
fin
d
in
g t
h
e c
h
aracte
r
istics, co-occ
urrence
matrix
is freq
u
en
t. Fro
m
th
e co
-o
ccurren
c
e
m
a
trix
set
of textural features extract
ed a
nd these fe
atures are
wid
e
ly u
s
ed
to
ex
tract tex
t
ural in
fo
rm
atio
n
fro
m
d
i
g
ita
l
i
m
ag
es [
7
],
[8].
I
n
stru
ct
ur
al ap
pr
o
a
ch
, textu
r
e is
co
nsid
ered
as
a rep
e
titio
n
of
so
m
e
p
r
i
m
itiv
e
s
. For tex
t
u
r
e
classificatio
n
an
d
ch
aracterizatio
n
,
these meth
od
s
have
bee
n
a
p
pl
ied by
seve
ral
authors
and
ac
hieve
d
s
u
ccess
to a ce
rtain
de
gree
[9].
C
h
aract
eri
zat
i
o
n an
d cl
assi
fi
c
a
t
i
on o
f
t
e
xt
ure
s
i
s
an i
m
port
a
nt
st
ep i
n
t
h
e st
udy
of
pat
t
e
rn
s
on t
e
xt
u
r
e
im
ages. The
textures are characterized
a
n
d cl
assi
fi
e
d
re
cent
l
y
by
va
ri
ous
pat
t
e
r
n
a
p
pr
oac
h
m
e
t
hod
s:
ed
g
e
di
rect
i
o
n m
o
v
e
m
e
nt
s [1
0]
, l
o
n
g
l
i
n
ea
r
pat
t
erns
[
11]
,
[
1
2
]
an
d
pre
p
r
o
ce
ssed i
m
ages [
13]
.
M
a
r
b
l
e
t
e
xt
u
r
e
descri
pt
i
on
[
1
4
]
, avoi
di
n
g
C
o
m
p
l
e
x Pat
t
e
rns [1
5]
, Te
xt
ure
im
ages are al
so de
scri
be
d a
n
d cl
assi
fi
ed
by
usi
n
g
vari
ous
wa
vel
e
t
t
r
ans
f
o
r
m
s
techni
que
s:
o
n
e
base
d
o
n
st
at
i
s
t
i
cal
param
e
ters
[1
6]
a
n
d
a
not
her
o
n
e
bas
e
d
o
n
p
r
im
it
iv
e p
a
ttern
s [1
7
]
.
Sasi
Ki
ra
n et
.
a
l
[1
8]
has
pr
op
ose
d
a m
e
t
hod cal
l
e
d
Wa
v
e
l
e
t
based
Hi
st
og
ram
on t
e
x
t
on
pat
t
e
rns
(
W
HPT
)
an
d g
r
o
u
p
ed t
h
e st
o
n
e t
e
xt
u
r
es i
n
t
o
fo
u
r
cat
ego
r
i
e
s. The
WHTP
m
e
t
hod g
o
t
av
erage % o
f
g
r
o
upi
n
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
S
t
on
e
Imag
e C
l
a
ssifica
tio
n Ba
sed
o
n
Overl
a
p
p
e
d
5
b
it T
Pa
ttern
s
Occu
rren
ce
o
n
…. (Pa
l
na
ti Vija
y K
u
ma
r)
1
153
as 94
.5
6.
D
r
.
U R
a
vi
B
a
b
u
e
t
.al
[1
9]
has p
r
op
ose
d
a m
e
t
hod
fo
r st
o
n
e t
e
xt
u
r
es cl
assi
fi
c
a
t
i
on i
n
t
o
f
o
u
r
gr
o
ups
.
In
t
h
i
s
m
e
t
hod
al
so
use
d
pat
t
erns
ap
pr
oac
h
o
n
grey
-t
o
-
g
r
ey
l
e
vel
p
r
e
p
r
o
cesse
d i
m
ages. T
h
i
s
m
e
t
h
o
d
al
s
o
achi
e
ve
d 9
7
.
1
5
%
as gr
o
up cl
a
ssi
fi
cat
i
on,
b
u
t
t
h
i
s
m
e
t
hod i
s
appl
i
e
d
o
n
l
y
f
o
r
gr
o
upi
ng st
o
n
e t
e
xt
u
r
es i
n
t
o
f
o
u
r
groups
. S
u
m
a
Latha et.al. [20] ha
s
proposed
a m
e
thod called LBP
-
High
-S
ym
m
e
try
(LBP
-HS
)
for rec
o
gnition
of
stone text
ures. T
h
is a
p
proach is also
patterned
ap
pr
o
ach f
o
r st
o
n
e
t
e
xt
ure
rec
o
g
n
i
t
i
on. T
h
e LB
P-H
S
m
e
t
hod
got
9
2
%
of rec
o
gni
t
i
on
onl
y
.
S
u
jat
h
a B
et
al
[21
]
pro
p
o
s
ed a
m
e
t
hod cal
l
e
d
Text
o
n
an
d T
e
xt
u
r
e
Ori
e
nt
at
i
on C
o
-
o
cc
ur
rence
M
a
t
r
i
x
(T&T
O-C
M
) f
o
r t
h
e cl
assi
fi
cat
i
on of t
e
xt
ure
s
.
The pr
o
p
o
s
e
d
m
e
t
hod
achi
e
ve
d onl
y
93
% of
cl
assi
fi
cat
i
on rat
e
.
In
m
o
st
app
r
oa
ches,
w
h
i
c
h
ha
ve
bee
n
of
fere
d s
o
far
,
resea
r
chers
have
t
r
i
e
d t
o
a
n
al
y
ze a
n
d
d
e
scri
be
t
e
xt
ure
base
d
on
o
v
erl
a
ppe
d
al
pha
bet
pat
t
e
rns
f
o
r
st
o
n
e
im
age cl
assi
fi
cat
i
on. T
h
e
p
r
o
p
o
sed
m
e
t
hod
put
fo
rwa
r
d t
h
e
pa
t
t
e
rn a
p
p
r
o
ach
fo
r
gr
o
upi
ng
t
h
e st
one
t
e
xt
u
r
es i
n
t
o
f
o
u
r
cl
asses. T
h
e
hi
g
h
acc
uracy
i
n
t
e
xt
u
r
e
classificatio
n
i
n
th
e resu
lts sh
ows th
e
qu
ality o
f
offered ap
pro
ach. Th
e
p
r
esen
t
p
a
p
e
r
p
r
op
o
s
es an
app
r
o
a
ch
fo
r st
one
t
e
xt
u
r
es cl
assi
fi
cat
i
o
n
base
d
on
occ
u
r
r
ence
s
of
o
v
e
rl
ap
ped
T-
pat
t
e
rns
o
n
eac
h
5
×
5 s
u
b
-
i
m
ages.
The r
e
m
i
nder
of t
h
i
s
pa
per
i
s
or
ga
ni
zed as
fol
l
o
ws:
Sect
i
on t
w
o
desc
ri
b
e
s t
o
t
h
e i
d
ent
i
fi
cat
i
on o
f
Ov
erlapp
ed
5-b
it T-Pattern
s
(O5TP) on
th
e
g
r
ey level i
m
ag
e.
Sectio
n th
ree is related
t
o
d
e
ri
v
i
ng
an
algo
rith
m
fo
r
gr
o
upi
ng
t
h
e st
o
n
e t
e
xt
ure
and analyses t
h
e res
u
lts and
fi
n
a
lly, th
e con
c
lu
sion
i
n
clud
ed.
2.
PROP
OSE
D
METHO
D
2.
1.
Identi
fi
c
a
ti
on
of
O
v
erl
ap
ped
5
-
bi
t T
-
P
a
t
t
e
r
ns o
n
E
a
ch
5
×
5 su
b-s
t
one i
m
a
g
e
The
p
r
o
p
o
sed
m
e
t
hod
O
5
TP
con
s
i
s
t
s
of
4 st
eps.
In
st
ep
1,
con
v
e
r
t
t
h
e eac
h st
o
n
e t
e
xt
ure
col
o
r i
n
t
o
the grey level im
age by using
7-bit
bina
ry code qua
n
tization
m
e
thod.
Ide
n
tify the 5-bit T-patterns i
n
each 5×
5
wi
n
d
o
w
of t
h
e
st
one t
e
xt
u
r
e i
m
age i
n
st
ep
2.
In
st
ep
3, c
o
u
n
t
t
h
e
occu
rre
n
ces of
T-
pat
t
e
r
n
s.
Fi
nal
l
y
, bas
e
d
o
n
t
h
e num
ber o
f
T-pat
t
e
r
n
s
deri
ve a ne
w al
go
r
i
t
h
m
for cl
assi
f
i
cat
i
on. The
bl
ock
di
ag
ram
of t
h
e ent
i
r
e pr
oc
edu
r
e
i
s
sh
ow
n i
n
Fi
g
u
re
1
.
Fi
gu
re
1.
B
l
oc
k
di
ag
ram
of t
h
e
proposed stone im
age classification
Step 1:
C
o
lor to
Grey Scale c
o
nve
r
sion
Th
e C
o
lor imag
e is
n
o
t
h
i
ng bu
t a co
lo
r ch
ann
e
ls.
Mo
st
d
i
g
ital im
ag
es are co
m
p
rised
o
f
t
h
ree
separate c
o
lor
channels: a
red cha
nnel,
a gre
e
n c
h
annel,
a
n
d a
blue c
h
a
n
nel.
Grey scale
means m
a
ny s
h
ade
s
(grey) fro
m
b
l
ack
to
wh
ite.
Gen
e
rally, 7
ways are av
aila
ble to conve
r
t the col
o
r im
age into gray scal
e im
age
i
.
e. ave
r
a
g
i
n
g
m
e
t
hod,
l
u
m
a
m
e
t
hod,
an
d
D
e
-sat
u
r
at
i
o
n
m
e
t
h
o
d
, C
u
st
om
#
o
f
g
r
ay
s
h
ad
es m
e
t
hod,
h
o
r
i
zont
al
error-di
ffusion
ditheri
ng m
e
thod, Sing
le color channel and Single col
o
r c
h
a
nnel m
e
thod.
In this pa
per
utilized
C
u
st
om
# o
f
gr
ay
sha
d
es m
e
t
hod
.
Cu
sto
m
#
of
g
r
ay sh
ad
es meth
od
: th
is allo
ws th
e
use
r
to s
p
ecify how m
a
ny shades
of
gray the
resulting im
age will use.
Thi
s
value c
a
n
be
betwee
n
2
and 256 is acce
pted. If it is
2,
the resultant image
cont
ai
n
s
2 s
h
a
d
es i
.
e
.
bl
ack
-a
nd
-w
hi
t
e
i
m
age, w
h
i
l
e
25
6
gi
ves a
n
i
m
age c
onsi
s
t
s
o
f
25
6
sha
d
es.
The
p
r
op
ose
d
m
e
t
hod
uses
8
-
bi
t
col
o
r c
h
a
n
n
e
l
s
. S
o
, m
a
xim
u
m
shades a
r
e
onl
y
25
6.
I
n
t
h
i
s
pa
per
uses
1
2
8
sh
ades
.
Any
grayscale conve
r
sion algor
ithm
is a thre
e-step
proces
s:
1.
C
a
t
c
h t
h
e
gr
ee
n,
re
d a
n
d
bl
ue
val
u
e
s
of a
pi
x
e
l
2.
C
o
n
v
e
r
t
th
o
s
e
th
r
e
e v
a
lu
e
s
in
t
o
a
sing
le gray
v
a
lu
e
3.
Replace the t
h
ree value
s
w
ith the
ne
w gray
value
Elaborated algorithm
for
Gre
y
Scale conve
rsion
St
ep
1:
E
x
cha
n
ge t
h
res
hol
d
va
l
u
e =
25
5/
(
N
u
m
ber Of
Sha
d
e
s
-1
)
Step
2: m
ean value =
(Red +
Gree
n +
Blue)
/ 3
St
ep
3:
G
r
ay
=
Int
e
ger
(m
ean val
u
e /
e
x
c
h
a
n
ge t
h
res
hol
d
va
l
u
e)
* e
x
c
h
an
g
e
t
h
res
h
ol
d
val
u
e
Step 2:
i
d
e
n
t
i
f
i
cat
i
on
of
5
-
bi
t
T-pat
t
e
r
n
s e
a
c
h
5×5
g
r
ey
-l
ev
el
st
one
su
b i
m
age
The
5×5 s
u
b i
m
age
values a
r
e represe
n
ted a
s
P
1
, P
2
, …
P
9
, P
10
, P
11
, … P
2
4,
P
25
. Th
e p
i
x
e
l p
o
s
ition
of
the each 5×
5 s
u
b wi
ndow
wa
s shown in
Figure
2.
Stone
t
e
xtu
re
color
Count the
T-p
a
tt
erns
U
s
er de
fi
ne
d
cl
as
si
fi
cat
ion
Cla
s
sif
i
ca
tion
Ide
n
tify
Ov
erlapp
ed
5
-
bi
t T
-
pa
t
t
er
ns
Grey
im
a
g
e
Color
Q
u
a
nt
izat
io
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E
V
o
l
.
6,
No
. 3,
J
u
ne 2
0
1
6
:
11
5
2
– 11
60
1
154
1 2 3
4 5
6 7 8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Fi
gu
re
2.
Pi
xel
p
o
si
t
i
ons
i
n
5
×
5
grey
l
e
vel
f
aci
al
sub
i
m
ag
e
In t
h
e
p
r
o
p
o
se
d m
e
t
hod co
ns
i
d
er t
h
e al
l
p
o
s
s
i
b
l
e
T-
pat
t
e
rn
fo
rm
ed usi
ng
5-
bi
t
s
. Th
e fi
rs
t
T-pat
t
e
r
n
fo
rm
ed usi
ng
5 pi
xel
P
1
, P
2
, P
3
, P
7
, and P
12
and t
h
e sec
o
n
d
T-pat
t
e
r
n
fo
r
m
ed usi
ng
5 pi
xel
P
2
, P
3
, P
4
, P
8
, and
P
13
and
so
o
n
.
From
fi
rst
r
o
w 3
T-
pat
t
e
rn
s are f
o
rm
ed.
From
t
h
e Sec
o
n
d
ro
w, a
n
ot
her
3 T
-
pat
t
e
r
n
s are
fo
rm
ed. From
the 3
rd
r
o
w a
n
ot
he
r set
of
3 pat
t
e
rns
fo
rm
ed. T
o
t
a
l
l
y
, ni
n
e
ove
rl
ap
ped
5-
bi
t
T-pat
t
e
r
n
s are
pos
sible in eac
h
5×5 s
u
b wi
ndow. Figure
3
shows
the
poss
ible overla
ppe
d T
-
pa
tterns on
each 5×
5 window.
Fi
gu
re
3.
O
v
erl
a
ppe
d
5
-
bi
t
T
-
pat
t
e
rns
o
n
ea
c
h
5×5
wi
nd
o
w
Step 3:
C
o
un
t
th
e T-p
a
ttern
s
Count the fre
quency occ
u
rre
nce of
the c
o
nsidere
d
patterns in each
5×5
sub
window on the stone
texture
adds t
h
ese val
u
es to the feature
vector.
Step 4:
Classification of st
one texture im
ages
Based on fre
quency occurre
nc
es of overla
pped 5-
bit T-patt
ern in each
5×
5 sub window
on the stone
texture
im
age is classified a
s
one
of t
h
e four
categ
ories i.e.
b
r
ick
,
Marb
le,
Mo
saic and
Gran
ite.
3.
RESULTS
A
N
D
DI
SC
US
S
I
ON
To find the effectivene
ss o
f
t
h
e pr
o
p
o
s
ed a
p
p
r
oach (
O
5T
P) carri
e
d
o
u
t
t
h
e expe
ri
m
e
nts on m
i
xed
sto
n
e
tex
t
ures
Dataset wh
ich
co
nsists of
v
a
rio
u
s
brick
,
gran
ite, m
a
rb
le, an
d m
o
saic sto
n
e
tex
t
u
r
es co
llected
fr
om
M
a
y
a
ng, Go
o
g
l
e
, C
u
R
e
t
Vi
sTex
, an
d Pa
ul
B
o
ur
ke
dat
a
-
b
ase a
n
d
al
so f
r
o
m
fl
oor i
m
ages t
a
k
e
n
by
ca
m
e
ra with t
h
e re
sol
u
tion
of
256×
256. T
h
e im
ages use
d
in
t
h
is ex
p
e
ri
m
e
n
t
is 1
8
8
0
i.e. 480
im
ag
es fro
m
M
a
y
a
ng
dat
a
b
a
se, 4
1
0
i
m
ag
es fr
om
Paul
B
o
u
r
ke
dat
a
ba
se, 160 im
ages from
VisTex
database
, 130
im
ages
fr
om
C
u
R
e
t
da
t
a
base
30
0 i
m
ages
fr
om
Go
og
l
e
dat
a
base
a
n
d
4
0
0
i
m
ages fr
om
scanne
d
ph
ot
o
g
r
ap
hs.
Fi
g
u
re
4
sh
ows th
e so
me o
f
th
e ston
e tex
t
ures u
s
ed
in th
is p
a
p
e
r to evaluate the efficiency
o
f
th
e
pr
opo
sed
ap
proach.
Som
e
of t
h
e f
r
e
que
ncy
o
f
oc
cur
r
ence
o
f
o
v
e
rl
ap
ped
5-
bi
t
T-pat
t
e
r
n
(
O
5
T
P)
of m
a
rbl
e
,
m
o
sai
c
, gra
n
i
t
e, an
d
bri
c
k t
e
xt
u
r
e
dat
a
set
im
ages are l
i
s
t
e
d-o
u
t
i
n
t
a
bl
e 1 t
o
4 res
p
ect
i
v
el
y
.
From
Tabl
es 1 t
o
5 des
i
gns a
cl
assi
fi
cat
i
on g
r
ap
h of
fi
ve
cat
eg
or
ies
o
f
stone i
m
ag
es is sh
ow
n in
Figu
r
e
5.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
S
t
on
e
Imag
e C
l
a
ssifica
tio
n Ba
sed
o
n
Overl
a
p
p
e
d
5
b
it T
Pa
ttern
s
Occu
rren
ce
o
n
…. (Pa
l
na
ti Vija
y K
u
ma
r)
1
155
Fi
gu
re
4.
som
e
o
f
t
h
e
st
o
n
e i
m
ages use
d
i
n
t
h
i
s
m
e
t
hod
wi
t
h
res
o
l
u
t
i
o
n
o
f
2
56×
2
5
6
Tabl
e 1. O
v
erl
a
ppe
d
5
-
bi
t
T-
pat
t
e
rns
occ
u
rr
ences of
B
r
i
c
k
t
e
xt
ures
I
m
age
Name
P
a
ttern1 P
a
ttern2
P
a
ttern3 P
a
ttern4 P
a
ttern5
Pattern6 Pattern7 Pattern8
Pattern9 Total
Br
ick.
0001
129
84
78
126
140
96
124
99
125
1001
Br
ick.
0002
122
138
136
122
133
129
137
135
128
1180
Br
ick.
0003
132
125
126
129
135
129
139
134
132
1181
Br
ick.
0004
133
147
131
136
155
128
137
141
124
1232
Br
ick.
0005
139
139
135
135
151
139
136
145
142
1261
Br
ick.
0006
140
140
142
139
147
138
143
151
139
1279
Br
ick.
0007
146
157
134
150
153
149
162
158
155
1364
Br
ick.
0008
170
175
156
168
168
158
168
168
158
1489
Br
ick.
0009
165
174
153
178
179
151
166
181
166
1513
Br
ick.
0010
168
171
158
168
172
161
176
181
164
1519
Br
ick.
0011
162
169
163
176
166
181
176
170
176
1539
Br
ick.
0012
212
220
185
199
205
180
196
224
199
1820
Br
ick.
0013
192
205
185
203
218
199
210
221
201
1834
Br
ick.
0014
232
249
245
202
218
217
201
218
216
1998
Br
ick.
0015
221
222
219
235
238
214
233
239
216
2037
Br
ick.
0016
251
256
239
259
277
253
279
299
267
2380
Br
ick.
0017
295
288
302
286
289
303
264
275
286
2588
Br
ick.
0018
429
433
416
424
431
402
418
421
406
3780
Br
ick.
0019
447
452
453
440
444
439
425
434
429
3963
Br
ick.
0020
616
631
438
584
412
468
608
613
618
4988
Tabl
e
2.
O
v
erl
a
ppe
d
5
-
bi
t
T-
pat
t
e
rns
occ
u
rr
ences
of
G
r
a
n
i
t
e
t
e
xt
u
r
es
Im
age Na
m
e
Pattern1
Pattern2
Pa
ttern3 Pattern4 Patter
n5 Patter
n6
Patter
n7
Pa
ttern8 Pattern9
Total
blue_gr
anite
2 3
1
2
1
5
4
0
0
18
blue_pear
l
2
1
2 1 2
3
1
4
4
20
blue_topaz
0
4
0 0 0
4
3
0
4
15
br
ick_er
o
sion
2
0
4 0 0
0
0
0
0
6
cany
on_black
0
1
4 2 0
1
0
0
1
9
dapple_gr
een
1
1
3 2 0
2
2
3
0
14
ebony
_oxide
0
1
2 1 2
4
0
1
0
11
giallo_gr
anite
1
4
0 0 0
0
2
0
1
8
gosf
o
r
d_sto
ne
0
0
2 3 4
2
1
1
0
13
gr
eenstone
0
1
4 0 2
3
2
0
0
12
interlude_haze
4
2
0 3 2
1
0
0
0
12
kalahar
i
0
4
2 2 1
1
3
1
1
15
m
e
sa_twilight
4
1
1 0 1
2
3
2
0
14
m
e
sa_ver
t
e
0
1
4 3 1
1
1
0
0
11
m
onza
4
0
0 3 2
1
0
1
0
11
pietr
o_ner
o
1
2
2 1 4
1
3
0
0
14
r
u
sset_gr
anite
0
4
1 1 0
0
1
0
0
7
gr
anite10
0
2
0 0 0
4
3
1
0
10
gr
anite13
4
3
2 0 1
1
0
1
0
12
gr
anite20
1
0
0 1 1
4
2
3
0
12
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E
V
o
l
.
6,
No
. 3,
J
u
ne 2
0
1
6
:
11
5
2
– 11
60
1
156
Tabl
e
3.
O
v
erl
a
ppe
d
5
-
bi
t
T-
pat
t
e
rns
occ
u
rr
ences i
n
Ho
ri
zont
al
Di
rect
i
o
n
o
f
M
a
r
b
l
e
t
e
xt
ures
Im
age Na
m
e
Pattern1
Pattern2
Pa
ttern3 Pattern4
Patter
n5
Patter
n6 Patter
n7 Pa
ttern8 Pattern9
Total
apollo
5
6
7 8
3
4 6
2
5
46
cany
on_blue
9
9
8
7
9
16
8
8
9
83
cotto 9
9
18
5
1
10
16
27
9
104
cur
r
y
_str
atos
8
9
8 9
1
6 9
9
8
67
flinder
s
_bl
ue
10
11
15
9
7
8 16
9
8
93
flinder
s
_gr
een
9
10
16
9
10
9
7
8
9 87
for
e
st_boa
10
11
10
16
10
8
8 10
8
91
for
e
st_sto
ne 10
10
10
10
11
9
11
16
10
97
gold
m
ar
ble1
10
10
16
11
11
9
11
9
7
94
gr
een_gr
anite
9
13
10
10
10
10
12
16
9
99
gr
ey
_stone
12
10
10
13
10
10
16
10
14
105
gr
ey
m
a
r
b
le1
9
10
12
11
10
10
10
11
9
92
gr
ey
m
a
r
b
le3
9
13
10
12
11
10
10
11
10
96
m
a
r
b
le001
11
12
11
11
12
10
16
10
10
103
m
a
r
b
le018
12
13
11
11
12
10
10
9
10
98
m
a
r
b
le034
11
15
13
18
9
16
11
13
17
123
m
a
r
b
le033 16
13
11
14
10
14
9
16
10
113
m
a
r
b
le012
10
14
11
10
33
30
15
15
12
150
m
a
r
b
le014
11
12
16
11
15
12
12
14
11
114
m
a
r
b
le020
15
15
13
15
13
12
10
12
11
116
Tabl
e
4.
O
v
erl
a
ppe
d
5
-
bi
t
T-
pat
t
e
rns
occ
u
rr
ences i
n
Ho
ri
zent
a
l
Di
rect
i
o
n
o
f
M
o
sai
c
t
e
xt
ures
I
m
age Na
m
e
Pattern
1
Pattern
2
Pattern
3
Pattern
4
Pattern
5
Pattern
6
Pattern
7
Pattern
8
Pattern
9
Total
concr
e
te_br
icks_1
7075
6
52
55
52
157
56
107
75
56
56
666
concr
e
te_br
icks_1
7075
7
58
56
52
155
57
106
73
56
55
668
concr
e
te_br
icks_1
7077
6
55
58
53
154
58
106
76
54
56
670
cr
azy
_paving_
509
1370
54
57
56
154
60
105
77
58
54
675
cr
azy
_paving_
509
1376
58
56
54
145
54
106
77
57
55
662
craz
y
_
tiles_1303
5
6
15
58
57
146
59
103
80
61
54
633
craz
y
_
tiles_5091
3
6
9
64
58
16
162
55
16
82
55
53
561
dirty_floor_tiles_footprin
ts_2564 95
62
59
156
59 106
79
61 60
737
dirty
_
tiles_20013
7
56
59
60
160
62
111
82
65
57
712
floor_tiles_030849
46
71
68
161 68
115
78
70 65
742
grubby
_tiles_25
65
58
66
65
165
67
112
88
76
72
769
kitchen_tiles_4270064
67 75
69
167
68
116
87
71 69
789
m
o
roccan_tiles_03
0826
71
74
79
165
63
121
83
65
70
791
m
o
roccan_tiles_03
0857
64
79
63
167
81
127
86
67
65
799
m
o
sai
c
_tiles_8071
010
71
70
72
173
68
122
91
72
71
810
m
o
sai
c
_tiles_leaf_pattern_201
005
06
0
66
76
74
165
76
123
90
74
75
819
m
o
sai
c
_tiles_ro
m
a
n_pattern_2
010
05
034
70
76
69
175
77
119
94
79
75
834
m
o
tif_tiles_61100
65
74
73
74
173
81
127
92
74
67
835
ornate_tiles_030
84
5
69
78
68
170
73
124
97
85
78
842
repeating_tiles_13
0359
80
74
64
184
81
115
107
91
63
859
Fi
gu
re
5.
The
pr
o
pose
d
m
e
t
hod
ge
ne
rat
e
d cl
assi
fi
cat
i
on
g
r
a
p
h
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
S
t
on
e
Imag
e C
l
a
ssifica
tio
n Ba
sed
o
n
Overl
a
p
p
e
d
5
b
it T
Pa
ttern
s
Occu
rren
ce
o
n
…. (Pa
l
na
ti Vija
y K
u
ma
r)
1
157
Fi
gu
re 6.
Ge
ne
rat
e
d
C
l
assi
fi
c
a
t
i
on Gra
p
h of m
a
rbl
e
an
d Gr
ani
t
e
st
o
n
e
t
e
xt
ures
The
gene
rat
e
d
gra
p
h sh
o
w
n i
n
fi
g
u
r
e 5
d
o
es
n’t
cl
earl
y
i
ndi
cat
e t
h
e gra
n
i
t
e
and m
a
rbl
e
be
cause
of t
h
e
occurre
nces
of 5-bit T
-
patterns a
r
e less c
o
m
p
are to ot
he
r tw
o gr
oup
s. So
, sep
a
r
a
te gr
ap
h is gen
e
r
a
ted
for
the
o
ccurren
ces of 5-b
it T-p
a
ttern
s i
n
m
a
rb
le an
d gran
ite
stone im
age. The
ge
nerate
d cla
ssification
gra
p
h for
m
a
rbl
e
an
d
gr
a
n
i
t
e
i
s
sh
o
w
n
i
n
fi
g
u
re
6.
F
r
o
m
t
h
e t
a
bl
es
1
t
o
4
an
d t
h
e cl
assi
fi
cat
i
on
gr
a
phs
o
f
Fi
g
u
re
5an
d
6
assign a
n
exact
and speci
fic classifica
t
i
on o
f
col
o
r st
one i
m
ages usi
ng rat
e
of rec
u
r
r
ences
of o
v
erl
a
ppe
d
5-
bi
t
T-pat
t
e
r
n
s.
A
new
al
g
o
ri
t
h
m
i
s
deri
ved
f
o
r cl
assi
fi
cat
i
o
n
am
ong t
h
ese
fo
u
r
cl
asses i
.
e.
Gra
n
i
t
e
, M
a
rbl
e
,
M
o
sai
c
, an
d B
r
i
c
k g
r
o
u
p
o
f
st
one t
e
xt
u
r
es b
a
sed o
n
t
h
e ab
ove t
a
bl
e v
a
l
u
e
s
and
gene
rat
e
d g
r
ap
h. The
r
a
t
e
o
f
occu
rre
nces
of
5-
bi
t
T-pat
t
e
r
n
s i
s
dep
e
n
d
e
n
t
on t
h
e di
m
e
nsi
on
of t
h
e t
e
xt
ure t
h
at
m
eans whe
n
di
m
e
nsi
ons
of
t
h
e i
m
age chan
ged;
t
h
e rat
e
o
f
occ
u
r
r
ence
s i
s
al
so c
h
an
ge
d.
To a
voi
d s
u
c
h
pr
o
b
l
e
m
s
t
h
e present
pa
per
de
ri
ve
d
a classificatio
n alg
o
rith
m
in
dep
e
nd
en
t
of t
h
e im
age size.
This algorithm categ
orizes the sto
n
e tex
t
u
r
es in
to
fo
ur
g
r
o
u
p
s i
r
respect
i
v
e
o
f
t
h
ei
r
di
m
e
nsi
o
n
s
. T
h
e
deri
ve
d
al
go
ri
t
h
m
uses 2
56×
2
5
6
di
m
e
nsi
o
n
as a
ben
c
h
mark
.
If th
e rat
e
of
o
ccurren
c
es of t
h
e test i
m
ag
e cat
aract
with
in
t
h
e
range of m
i
n
i
m
u
m
to
m
a
x
i
m
u
m
q
u
a
n
tity
o
f
o
c
cu
rren
ces o
f
two
an
d
fou
r
tran
sition
s
of a fastid
i
o
u
s
sto
n
e
th
en
test imag
e is categ
orized
as a
p
a
rti
c
u
l
ar
g
r
ou
p.
A
l
go
rit
h
m 1:
St
one
t
e
xt
u
r
e c
l
assi
fi
cat
i
on
ba
sed
o
n
O
v
e
r
l
a
p
p
ed
5
-
bi
t
T-Pat
t
erns
Let
Su
m
o
f
o
c
cu
rren
ces o
f
Overlapp
ed
5-b
it T-Pattern
s
SOTP
START
if SOTP<=
t
h
en
Test im
age texture
group is c
a
tegorized a
s
GRANIT
E cla
ss
Othe
rwise i
f
S
O
TP <
=
th
en
Test im
age texture
group is c
a
tegorized a
s
MARBLE clas
s
Othe
rwise i
f
S
O
TP <
=
th
en
Test im
age texture
group is
c
a
tegorized a
s
MOSAIC class
Othe
rwise i
f
S
O
TP<=
(
th
en
Test im
age texture
group is c
a
tegorized a
s
BRICK class
Othe
rwise
Test im
age texture
group is
categ
o
r
i
zed
as
U
NKN
OW
N class
STOP
4.
COMPARISON BETWEE
N PROPOSE
D
METHOD
AND OTHER
E
X
ISTING METHODS
The
pr
op
ose
d
m
e
t
hod i
s
c
o
m
p
are
d
wi
t
h
W
a
vel
e
t
base
d Hi
s
t
og
ram
on Tex
t
on Pat
t
e
r
n
s
(
W
H
T
P)
[
18]
,
whi
c
h i
s
use
d
t
o
cl
assi
fy
t
h
e s
t
one t
e
xt
ure i
m
ages i
n
t
o
f
o
u
r
cat
ego
r
i
e
s by
usi
n
g wa
vel
e
t
based t
e
xt
o
n
pat
t
e
r
n
hi
st
o
g
ram
and
t
e
xt
on
feat
u
r
e
evol
ut
i
o
n m
e
tho
d
[2
2]
, w
h
i
c
h i
s
use
d
t
o
cl
assi
fy
t
h
e i
m
ages i
n
t
o
f
o
ur
g
r
o
u
p
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E
V
o
l
.
6,
No
. 3,
J
u
ne 2
0
1
6
:
11
5
2
– 11
60
1
158
base
d o
n
rat
e
of
occ
u
r
r
ence
s
of t
e
xt
o
n
p
a
t
t
e
rns
.
T
h
e p
r
op
ose
d
m
e
t
hod i
s
al
so c
o
m
p
ared wi
t
h
ot
her e
x
i
s
t
i
n
g
m
e
t
hod l
i
k
e S
y
nt
act
i
c
Pat
t
e
r
n
o
n
3D m
e
t
hod [
2
3]
i
n
whi
c
h st
one t
e
xt
ur
es are cl
assi
fi
ed i
n
t
o
f
o
u
r
cat
ego
r
i
e
s
base
d o
n
t
h
e
o
ccur
r
ence
o
f
s
y
st
em
ati
c
pat
t
e
rns
.
It
i
s
cl
earl
y
ob
vi
o
u
s t
h
at
,
t
h
e p
r
o
p
o
se
d
m
e
t
hod s
h
ow
s
i
gns
of
a h
i
gh
classificatio
n
rate th
an
th
e ex
isting
meth
od
s. Th
e
pe
rcenta
ge m
ean classifica
t
i
on r
a
t
e
for t
h
e
pr
o
pos
ed
m
e
t
hod a
n
d
ot
her
exi
s
t
i
n
g m
e
t
h
o
d
s a
r
e
rep
r
esent
e
d
i
n
Tabl
e 5.
T
h
e
gra
phi
cal
rep
r
ese
n
t
a
t
i
o
n
o
f
t
h
e
perce
n
t
a
ge
m
ean cl
assi
fi
cat
i
on rat
e
f
o
r t
h
e p
r
o
p
o
sed
m
e
t
hod an
d ot
her e
x
i
s
t
i
ng m
e
t
h
o
d
s are s
h
o
w
n i
n
Fi
gu
re
7. T
h
e
Tabl
e 5 a
n
d Fi
gu
re 7
sh
o
w
s t
h
e m
ean perce
n
t
a
ge cl
assi
fi
ca
t
i
on o
f
o
r
i
g
i
n
al
im
ages Go
ogl
e and sca
n
ne
d im
age
.
The m
ean percent
a
ge cl
assi
fi
cat
i
on o
f
pr
o
p
o
se
d
m
e
t
hod a
nd
ot
he
r exi
s
t
i
ng m
e
t
hods o
f
vari
ous
dat
a
ba
ses are
rep
r
ese
n
t
e
d i
n
Tabl
e
6 a
n
d
g
r
aphi
cal
rep
r
ese
n
t
a
t
i
on i
s
sh
o
w
n i
n
Fi
g
u
r
e
8.
Table 5.
Mea
n
perce
n
tage
clas
sification res
u
lts
of the
proposed m
e
thod an
d
othe
r e
x
isting m
e
thods
I
m
age Dataset
Syntactic
Pattern o
n
3D
m
e
thod
W
a
velet based
His
t
ogr
am
on T
e
xton Patter
n
s
T
e
xton Featur
e
Detection
Pr
oposed
M
e
thod
Or
iginal 93.
29
93.
15
95.
56
96.
85
Google
92.
53
92.
87
94.
15
96.
35
Scanned 93.
3
93.
82
95.
27
96.
29
Aver
age 93.
59
93.
28
94.
97
96.
19
Fi
gu
re
7.
C
l
assi
fi
cat
i
on c
h
art
of
p
r
o
p
o
se
d m
e
t
h
o
d
wi
t
h
ot
h
e
r e
x
i
s
t
i
ng m
e
t
h
o
d
s
Tabl
e 6.
M
e
a
n
perce
n
t
a
ge
cl
as
si
fi
cat
i
on rat
e
s of
t
h
e pr
o
p
o
s
e
d
m
e
t
hod
a
n
d ot
he
r
e
x
i
s
t
i
n
g
m
e
t
hods
I
m
age Dataset
Syntactic Pattern
o
n
3D m
e
thod
W
a
velet based
His
t
ogr
am
on T
e
xton Patter
n
s
T
e
xton Featur
e
Detection
Pr
oposed
M
e
thod
VisT
ex 93.
15
92.
87
95.
46
95.
95
T
e
xtur
e I
m
ages
T
a
ken by
Ca
m
e
ra
92.
87
91.
7
95.
12
96.
35
CuReT
93.
32
93.
56
94.
86
96.
76
M
a
yang 92.
83
92.
95
94.
39
95.
85
Paul Bour
ke
93.
05
93.
05
95.
23
95.
93
Fi
gu
re
8.
M
ean
pe
rcent
a
ge cl
a
ssi
fi
cat
i
on c
h
a
r
t
of
t
h
e
pr
o
p
o
s
ed m
e
t
hod a
n
d
ot
he
r e
x
i
s
t
i
n
g
m
e
t
hods
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
S
t
on
e
Imag
e C
l
a
ssifica
tio
n Ba
sed
o
n
Overl
a
p
p
e
d
5
b
it T
Pa
ttern
s
Occu
rren
ce
o
n
…. (Pa
l
na
ti Vija
y K
u
ma
r)
1
159
No st
a
nda
r
d
cl
assi
fi
cat
i
on al
go
ri
t
h
m
i
s
used t
o
t
e
st
t
h
e dat
a
base. T
h
e no
vel
t
y
of t
h
e p
r
o
p
o
s
e
d
m
e
t
hod i
s
t
h
at
t
h
e pr
o
p
o
s
ed t
echni
que i
s
a
p
pl
i
e
d o
n
hu
ge
dat
a
set
.
Eve
n
t
h
o
u
gh i
t
i
s
ap
p
l
i
e
d on
h
uge
d
a
t
a
set
it
gives
good res
u
lts whe
n
c
o
m
p
are
with the
othe
r exis
ting
m
e
thods
. Still, no s
u
c
h
technique is available to
appl
y
on
l
a
r
g
e dat
a
set
.
5.
CO
NCL
USI
O
N
The p
r
ese
n
t
p
a
per
deri
ved a
new a
p
pr
oac
h
cal
l
e
d O
v
e
r
l
a
ppe
d
5-
bi
t
T-
Pat
t
e
rns
(O
5T
P) f
o
r st
o
n
e
texture classifi
cation. The present
pape
r conside
r
ed
Nine 5-bit T-pa
tterns on each 5×
5 sub im
age without
lo
sing
th
e in
fo
rm
atio
n
ab
out th
e i
m
age f
o
r text
ure ana
l
ysis of the gr
ey lev
e
l i
m
ag
e.Th
e
n
o
v
e
lty o
f
th
e
pr
o
pose
d
m
e
t
hod i
s
n
o
st
an
da
rd cl
assi
fi
cat
i
o
n al
g
o
ri
t
h
m
i
s
use
d
f
o
r cl
assi
f
i
cat
i
on o
f
st
o
n
e
t
e
xt
ures
. P
r
o
pos
ed
m
e
t
hod i
s
t
e
st
ed by
usi
n
g l
a
rge set
dat
a
ba
se and
g
o
t
hi
g
h
% o
f
g
r
ou
p
cl
assi
fi
cat
i
on i
.
e. t
h
e st
re
ngt
h
of t
h
e
propose
d
m
e
thod.
When c
o
m
p
are with
the othe
r
exi
s
ting m
e
thod gives m
o
re accurate a
nd
precise
classificatio
n
resu
lts. The O5TP is co
m
p
u
t
atio
n
a
lly in
ex
pe
nsive
.
The e
xperim
e
ntal results clearly indicate the
efficacy of t
h
e
propose
d
O5T
P
ov
e
r
the
va
rious
e
x
isting m
e
thods.
REFERE
NC
ES
[1]
C. H. Chen,
et
al
.
, “Handbook of
Pattern Recognition and Computer
Vision,” 2nd
ed. Singapor
e,
World Scientific,
2000.
[2]
R. M. Haralick
,
et al
.
,
“
T
extur
a
l featur
es
for i
m
a
ge clas
s
i
fi
cat
ion,”
I
EEE T
r
ans. Syst., Man
,
Cybern
., vo
l/iss
u
e:
3(6), pp
. 610–62
1, 1973
.
[3]
F. S. Cohen,
et al.
, “Automated inspection of
textile
fab
r
ics u
s
ing textural models,”
I
EEE T
r
ans. Pattern Ana
l
.
Mach. Intell
., vo
l/issue: 13
(8), pp
. 803–808
, 1991
.
[4]
R. N. Sutton and E. L. Hall, “Texture measures
for autom
a
tics
clas
s
i
fi
cat
i
on of pulmonar
y
disease,”
IEEE Trans.
Comput
., vo
l/iss
u
e: C-21(7)
, pp
.
667–676, 1972
.
[5]
H.
Harms,
et a
l
.
,
“Combined
local color and textur
e an
al
ys
is
of s
t
ain
e
d c
e
ll
s
,
”
Comput.
Vis
.
Graph., Imag
e
Proc
e
ss.
, vol/issue: 33(3)
, pp
. 36
4–376, 1986
.
[6]
N.
Hamdi,
et
a
l
., “A new appr
oach B
a
sed on
Quantum Cluste
ring and
Wavelet
Transform fo
r breast cancer
Clas
s
i
fic
a
tion
:
Com
p
arative
s
t
ud
y,”
In
ternatio
nal Journal of
Electrica
l
and
Computer Engineering (
I
JECE)
,
vol/issue:
5(5), p
p
. 1027-1034
, 2
015.
[7]
T. Chang and C
.
C. J
.
Kuo, “
T
exture an
al
ys
is
and clas
s
i
fi
cat
io
n with tree-s
t
ru
ctured wave
let t
r
ans
f
orm
,
”
IEEE
Trans. Image Pr
ocessing
, vo
l/iss
u
e: 2(4)
, pp
. 429
-442, 1993
.
[8]
J. L. Chen
and
Kundu, “Unsupervised t
e
xture segm
enta
tion
using m
u
lti-ch
annel
decom
position
and hidden M
a
rk
ov
models,”
I
EEE Trans. Image.
Processing
, vo
l/issu
e: 4(5)
, pp
. 603-
620, 1995
.
[9]
R. W. Conners,
“Toward a set o
f
statis
tical features which measur
e visually
p
e
rceiv
a
ble qu
alities of textur
e,” in
Proc.
Patt
ern R
e
cognition
Image
Processing Conf
.
, pp. 382-390, 1
979.
[10]
B. E. R
e
dd
y
,
et al.
, “
T
exture
Cl
assifica
tion b
y
s
i
m
p
le pa
tterns o
n
edge dir
e
ction
movements,”
IJCSNS
, vol/issue:
7(11), pp
. 220-2
25, 2007
.
[11]
V.
V.
Krishna,
et al.
, “Classification of textur
es based on distan
ce
function of linear pattern
s using mathematical
morpholog
y
,
”
P
r
oceedings o
f
IC
EM,
conduc
ted
b
y
JNT Uni
versit
y
, India
, 2005
.
[12]
V.
V.
Ku
mar,
et al.
, “An Innova
tive Techniqu
e of Texture Class
i
fication and Co
mparison Based on Long Linear
Patterns,
”
Journal of Computer S
c
ien
c
e,
vol/issue: 3(8), pp. 633-6
38, 2007
.
[13]
V.
V.
Ku
mar,
et al.
, “A
measure of patterns tren
ds on va
rious ty
pes of preprocessed images,”
IJCSNS
, vol/issue:
7(8), pp
. 253-25
7, 2007
.
[14]
A.
Suresh,
et a
l
.
, “An Innovative Techniqu
e
of Marble Text
ure Description
Based on Grain Components,
”
International Jo
urnal of Comput
er Science and
N
e
twork S
ecurity
,
vol/issue:
8(2), p
p
. 122-126
, 200
8.
[15]
V.
V.
Kumar,
et
al
., “Classification of Textur
es b
y
Avoiding
Co
mplex Patterns,
Science publications,”
Journal o
f
Computer Scien
c
e
, 2008.
[16]
U.
S.
N. Raju,
et al.
,
“
T
extur
e
Description
usin
g Differe
nt W
a
v
e
le
t
Tr
ansforms Based on Sta
tist
i
ca
l Param
e
ters,
”
proceedings of
t
h
e 2nd WSEAS I
n
ternational S
y
mpos
ium on WAVELETS THEOR
Y
&
APPLICA
T
IONS in Applie
d
Mathematics, S
i
gnal Processing
&
M
odern
Scien
ce (
W
AV '08)
, Istanbul, Turkey
, p
p
. 174-178
, 200
8.
[17]
V.
V.
Kumar,
et al.
, “A New Method of Texture Cl
assification using var
i
o
u
s Wa
velet
Transforms based on
Prim
itive Patter
n
s,”
ICGST International Journal on Graphi
cs, Vision and Image Processing, GVIP
, vol/issue
:
8(2), pp
. 21-27
,
2008.
[18]
J.
S.
Kira
n,
et a
l
.
, “Wavelet based
Histogram meth
od fo
r classification
of textur
es,”
IJECT
, vo
l/issue: 4(3), pp. 149-
164, 2013
.
[19]
U. R. Babu
,
et
al
.,
“T
e
x
t
u
re
Cl
a
ssi
fi
ca
t
i
on ba
sed on Tex
t
on
Patterns using
on
various Grey
to Grey
lev
e
l
Preprocessing Methods,”
In
ter
national
Journal
of Signa
l
Processi
ng, Image Pr
ocessing and
Pa
ttern R
ecogn
itio
n
,
vol/issue: 6(4),
2
013.
[20]
L. Sum
a
la
tha
a
nd B. Suja
tha
,
“
A
New Approach for
R
ecogn
ition of Mosaic
Textur
es b
y
LB
P Based On R
G
B
Model,”
Signal
&
I
m
age Processing: An
In
terna
tional Journal
(
S
IPIJ)
,
vol/issue:
4(1), 2013
.
[21]
B. Sujath
a,
et
al
., “Textur
e
Classification Using
Texton Co-Occurr
e
nce Matrix
Der
i
ved From Textu
r
e Orien
t
ation,”
International Jo
urnal of So
ft Co
mputi
ng and
En
gineering
(
I
JSCE)
ISSN: 2231-2
307
, vol/issue: 2
(
6), 2013
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E
V
o
l
.
6,
No
. 3,
J
u
ne 2
0
1
6
:
11
5
2
– 11
60
1
160
[22]
U. R. Babu,
et a
l
.
, “Tex
ture Clas
sification Based
on Texton Featu
r
es,”
I.J. Image, Graphics and Si
gnal Processing
,
vol/issue:
4(8), p
p
. 36-42
.
[23]
Suresh A.
and V.
V.
Kumar,
“P
attern Based C
l
assification of
St
one Textur
es on a Cubical Mask,”
Internatio
nal
Journal of Universal Computer S
c
ien
ces
, vol/issue” 1(1)
, pp
. 4
-
9,
2010.
BIOGRAP
HI
ES OF
AUTH
ORS
Palnati Vijay
Kumar
received his B.Tech (Computer Science & Engineering) from
SRKR
Engineering College. He is pur
suing his M.Tech
(CSE) from Ad
it
y
a
College of
Engineering,
Suram
p
alem
, aff
ilia
ted to JNT Universit
y
Kaki
na
da, Kak
i
nada
.
His research in
terests inc
l
ude
image processin
g
,
cloud computing.
Pullela
S V V S R Kum
a
r is wo
rking as Profess
o
r of CSE
at A
d
it
ya Co
lleg
e
of
Engin
eering
,
S
u
ram
p
alem
. He
rece
ived h
i
s
Doctora
t
e from
Ac
ahr
y
a Naga
rjuna
Univers
i
t
y
, An
dhra P
r
ades
h.
He is having more than 16
y
ears of experience
and published 12 research pap
e
rs in various
International Jo
urnals and
Conf
erences. His
re
s
earch
int
e
res
t
s
i
n
clude
Data Mining, Pattern
Recognition an
d Image Processing. He acted
as a reviewer to va
rious
International
Conferenc
e
s
.
Nakkella
Madhu
ri
is working as A
ssistant Profe
ssor of CSE
at Adity
a
College of
En
gineer
ing,
S
u
ram
p
alem
.. S
h
e rec
e
iv
ed her
M
.
Tech
(Com
puter S
c
i
e
nc
e
& Engin
eering)
from
J
N
T
U
Kakinada, Kak
i
n
a
da. Her
re
se
arch inte
re
sts inc
l
ude
ima
g
e
processing and
cloud
computing.
M
.
Um
a
Devi is
working as As
s
o
ciate P
r
ofes
s
o
r of CS
E at Adit
ya Coll
ege of
Engineerin
g,
S
u
ram
p
alem
.. S
h
e rec
e
iv
ed her
M
.
Tech
(Com
puter S
c
i
e
nc
e
& Engin
eering)
from
J
N
T
U
Kakinada, Kakinada and pursu
ing her Ph.D.
from Achar
y
a N
a
garjun
a Univer
sity
, Guntur,
Andhra Pradesh. Her r
e
sear
ch in
terests in
clude im
age pro
cessing
and cloud
computing.
Evaluation Warning : The document was created with Spire.PDF for Python.