Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
10
,
No.
4
,
A
ugus
t
2020
,
pp.
3550
~
35
57
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
10
i
4
.
pp
3550
-
35
57
3550
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om/i
nd
ex
.ph
p/IJ
ECE
Text hid
ing in te
xt
usin
g
invisib
le cha
racte
r
Nada A
bdul
Az
i
z
M
usta
fa
Coll
ege of
La
ng
uage
s,
Univ
ersity
of
Baghd
ad, I
r
aq
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
un
24
,
2019
Re
vised
Jan
2
9
,
20
20
Accepte
d
Fe
b 10
, 202
0
Stega
nogra
ph
y
c
an
be
def
in
ed
as
the
art
and
sc
ience
of
hidi
ng
inf
orm
at
ion
in
the
data
that
c
ould
be
rea
d
b
y
computer
.
Th
is
scie
nce
ca
nn
ot
rec
ogni
ze
stego
-
cove
r
and
the
orig
ina
l
one
whethe
r
b
y
e
y
e
or
b
y
computer
when
see
in
g
the
sta
ti
sti
cal
sa
m
ple
s.
Thi
s
p
ap
er
pre
sen
ts
a
n
e
w
m
et
hod
to
hid
e
t
ext
in
t
ext
cha
ra
cters.
Th
e
s
y
stemat
ic
m
e
tho
d
uses
the
stru
cture
of
inv
isibl
e
cha
ra
cter
to
hide
and
ext
r
act
sec
ret
t
ext
s.
Th
e
creat
ion
of
sec
ret
m
essage
co
m
prises
four
m
ai
n
stage
s
su
ch
using
the
l
et
t
er
from
t
h
e
origi
nal
m
essage
,
se
lecti
n
g
the
suit
abl
e
cov
er
te
x
t,
d
ivi
ding
the
cove
r
te
xt
i
nto
bloc
ks,
hiding
the
se
cret
te
xt
using
the
i
nvisibl
e
ch
aracter
and
compari
ng
th
e
cove
r
-
text
and
stego
-
obje
c
t.
Thi
s
stu
d
y
uses
an
invis
ibl
e
cha
ra
ct
er
(
white
spa
ce
)
po
siti
on
of
in
the
cove
r
t
ext
tha
t
used
to
hide
t
he
the
sec
ret
e
sende
r
m
assege
s.
The
exp
eri
m
ents
result
s
show
tha
t
the
suggest
ed
m
et
hod
pre
s
ent
s
hig
h
l
y
sec
ret due t
o
use
t
he
m
ult
i
-
l
evel of c
om
ple
xi
t
y
to
avoi
d
th
e
a
ttack
ers.
Ke
yw
or
d
s
:
Invisible c
har
a
ct
er
Stegan
ogra
phy
Stego
-
c
ov
e
r
Stego
-
obj
ect
Stego
-
te
xt
Copyright
©
202
0
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Nad
a
Ab
du
l
A
zi
z Mustafa,
Coll
ege
of
Lan
gu
a
ges
,
Un
i
ver
sit
y o
f B
aghda
d
,
Karr
ad
a,
Al
-
J
ad
riya,
Ba
ghdad
,
Iraq
.
Em
a
il
:
nad
a@
colan
g.uoba
ghdad.ed
u.
i
q
1.
INTROD
U
CTION
Text
ste
gano
graph
y
te
ch
nique
is
app
li
ed
for
savin
g
pri
vacy
and
ori
gina
li
ty
of
te
xt
-
bas
ed
do
c
um
ents.
Ther
e
by,
te
xt
s
te
ganogra
ph
y
connsi
der
s
a
s
a
chall
en
ging
m
issi
on
that
te
nuous
a
dju
stm
ent
in
te
xt
file
can
be
sp
eci
fied
.
Invis
ible
cha
racter
t
echn
i
qu
e
is
use
d
to
hid
e
t
he
t
ext
int
o
te
xt
or
i
m
age
with
out
anyo
ne
can
be
seen
the
hidde
n
pro
cessi
ng
[1]
.
T
he
re
are
dif
fer
e
nt
te
chn
i
qu
es
l
ike
ste
gano
gr
a
ph
y,
c
rypto
gr
a
ph
y,
c
od
i
ng,
et
c
have
been
util
iz
ed.
Curr
e
ntly
,
ste
gano
gr
a
phy
is
com
m
on
ly
us
e
d
to
hi
de
te
xt
in
te
xt.
S
te
ganogra
ph
y
i
m
plies
“cov
e
re
d wr
it
ing
”
in G
reek
a
s w
el
l as it
c
onsiders
the
flag of
co
m
m
un
ic
at
ing
i
n
a
hidde
n st
yl
e
[2]
.
In
view
of
thi
s
dig
it
al
ste
ga
nograp
hy
co
nc
eal
s
eve
n
the
evide
nce
of
encr
y
pted
m
essagin
g
[
3
].
Stegan
ogra
phy
is
one
of
t
he
ol
dest
arts
that
pe
op
le
we
re
ea
ge
r
to
ha
ve
si
nc
e
they
sta
rte
d
c
omm
un
ic
at
ion
with
each
oth
e
r
[
4
]
.
Ma
ny
m
et
hods
have
been
us
e
d
to
hi
de
i
nfor
m
at
ion
by
us
in
g
t
he
rec
order
with
ta
le
s
of
ste
gano
gr
a
ph
y
and
c
rypto
gr
a
ph
y
th
rou
gh
ti
m
es
of
war
or
peace
[
5
]
.
Mor
eov
e
r,
ste
gano
gr
a
phy
is
the
art
an
d
sci
ence
w
hich
hid
e
inf
orm
ation
in
a
ny
co
m
pu
te
r
read
a
bl
e
data
in
a
way
that
a
ste
go
-
ob
j
ect
shou
l
d
be
no
t
disti
nguish
a
ble
from
or
igi
n
c
ov
e
r
neithe
r
by
a
hum
an
nor
by
com
pu
te
r
lookin
g
f
or
sta
ti
sti
cal
patte
rn
[
6]
.
Wh
il
e
ste
ga
no
gr
a
phy
is
an
ea
rly
top
ic
,
the
c
urren
t
pr
e
pa
rati
on
of
it
ar
rive
s
via
the
pr
is
oner
’s
iss
ues
s
ugge
ste
d
by
[
7
]
,
Alic
e
and
B
ob
a
re
the
two
pr
is
on
er
s,
they
com
m
un
ic
at
e
via
secret
to
acce
s
s
an
escape
st
rategy.
Seve
ral
m
essa
ge
se
nt
via
a
warde
n
cal
le
d
Eve
that
m
ay
t
os
s
the
pri
vate
lim
it
at
ion
when
it
su
s
pects
any
ty
pe
of
secre
t
m
essage
[
8
]
.
Th
us
,
the
sec
ret
le
tt
er
pr
ov
i
des
t
he
natu
ral
of
st
egano
gr
a
phy
to
re
al
iz
e
the
hid
in
g
inf
or
m
at
ion
be
tween
them
.
They
us
e
d
a
wa
rd
e
n
m
et
ho
d
t
o
te
st
the
w
hole
m
essages
re
placed
betwee
n
them
m
ay
e
it
her
be
direct
or
in
dire
ct
.
A
direct
wa
rd
e
n
m
et
ho
d
at
tem
pts
fo
r
ad
ju
sti
ng
the
m
essa
ge
an
d
the
sup
po
s
ed
cov
e
re
d
data
c
al
culat
ing
ly
vi
a
el
i
m
inati
ng
t
he
data
wh
il
e
an
i
nd
irect
wa
rd
e
n
m
et
ho
d
t
hat
is
us
e
d
to
ta
ke
the
m
eaning
o
f
cov
e
red
m
essage,
up
dates
the
oth
ers
a
nd
le
ts
the
m
essage
to
pe
rm
i
t
[
9
]
.
A
su
pp
os
it
ion
m
ay
be
com
plete
d
based
on
this
ty
pical
wh
en
bo
t
h
the
send
e
r
an
d
receive
r
will
sh
are
pa
rtic
ular
public
secre
t
data
then
the
c
onf
orm
ing
ste
gano
gr
a
phy
proce
dure
is
rec
ogni
zed
as
the
n
th
e
secret
key
s
te
ganogra
ph
y
wh
e
re
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Text
h
idi
ng in
text
u
sin
g
i
nvisible
char
acter
(
Nada A
bdul
Az
iz Must
afa
)
3551
as
pure
ste
ga
nograp
hy
pro
ceeds
that
the
re
is
nothin
g
pr
e
vious
data
m
utu
al
by
sen
der
a
nd
re
cei
ver.
Wh
il
e
the
publ
ic
key
of
th
e
re
cei
ver
is
ide
ntifie
d
to
t
he
sen
der,
the
ste
ga
nogra
ph
ic
proce
dure
is
nam
ed
public
key stega
nogra
ph
y
[
10
].
2.
R
EL
ATED
W
ORK
Stegan
ogra
phy
suppose
the
opposit
io
n
m
ay
ca
pture
the
co
ver,
ne
ve
rthele
ss
m
ay
be
no
t
ob
s
er
ve
a
ny
data
f
ur
t
her
t
he
or
i
gin
al
c
over
sat
isfie
d
[
11
]
.
T
he
data
i
s
co
ver
e
d
a
nd
can
hav
e
no
su
pple
m
entary
safety
furthe
r
the
r
ea
l
le
tt
er
e
m
bed
ding
[
1
2
]
.
A
r
ole
pri
nciple
f
or
perf
or
m
ance
m
easur
em
ent
of
a
ste
ga
no
g
ra
phic
structu
re
is
th
e
sta
ti
sti
cal
hi
dd
e
nness
of
t
he
sec
ret
data
[
13
]
.
Digital
te
xt
i
m
age
ha
ve
higher
grade
of
redu
nd
a
ncy
a
nd
hen
c
e
the
m
os
t
a
pprop
riat
e
f
or
ste
ga
nogr
aph
y.
T
her
e
e
xist
m
any
ste
gan
og
raphic
m
e
thods
that
al
low
hid
i
ng
sec
ret
data
in
te
xt
i
m
age
hav
in
g
their
ow
n
pros
an
d
co
ns.
Least
Sign
ifi
cant
Bi
t
m
od
ifi
cat
ion
(LSB)
m
et
ho
d
is
the
com
m
on
ly
us
ed
ste
ga
nogra
ph
ic
proce
dure
that
us
es
the
inf
or
m
at
ion
at
le
ast
sign
i
ficant
bits
of
a
te
xt
i
m
age
are
ra
nd
om
no
ise
an
d
s
hiftin
g
them
do
es n
ot
s
hift
th
e
per
ce
pt
ual wort
h
of
the
te
xt
i
m
ages.
More
ov
e
r,
LS
B
de
pend
on
a
ppr
oach
es
m
ay
be
ei
t
her
LSB
Ma
tc
hing
or
LSB
Re
place
m
ent
wh
e
re
t
he
f
or
m
er
swaps
the
L
S
B
of
t
he
pi
xels
with
the
co
m
m
un
ic
at
ion
to
be
sent
a
nd
the
ad
va
nced
increm
ent/decrem
ent
the
pi
xels
a
rb
it
rar
il
y
with
secr
et
bits
[
1
4
]
.
Hi
ding
in
f
or
m
at
i
on
withi
n
plac
es
ap
pears
to
be
pote
ntial
as
hum
an
hardly
m
a
y
disti
ng
uis
h
a
bout
the
bein
g
of
t
he
hidden
bits.
Pr
ese
ntly
,
op
e
r
at
ion
of
wh
it
es
paces
a
pp
ea
rs
us
ef
ul
and
has
it
s
po
s
sible
in
inf
or
m
at
ion
hi
ding
due
to
w
hitespa
ces
see
m
in
a
do
c
um
ent
m
or
e
than
the
occ
urrenc
e
of
wor
ds
.
It
is
un
i
form
a
ben
efit
wh
e
n
any
one
will
disti
nguish
that
a
blank
piece
of
te
xt
is
reall
y
ro
le
secret
inf
or
m
at
ion
[
1
5
]
.
I
n
a
ddit
ion
,
a
ne
w
te
c
hn
i
qu
e
of
te
xt
ste
gano
gr
a
phy
was
prese
nted.
T
heir
propose
d
a
lgorit
hm
us
es
te
xt
m
edia
to
e
m
bed
thei
r
se
cret
te
xt
file
de
pendin
g
on
a
dicti
on
a
ry.
T
hi
s
dicti
onary
c
onta
ins
En
glish.
Wo
r
ds
sorte
d
in
al
phabeti
cal
orde
r
to
be
sel
ect
e
d
by
us
er
in
orde
r
t
o
bu
il
d
t
he
c
over
te
xt
[1
6
]
.
The
sta
ti
sti
cal
pr
ope
rtie
s
of
dithered
i
m
a
ger
y
was
use
d.
W
it
h
this
m
et
ho
d,
the
i
nfor
m
at
ion
bits
to
be
con
ceal
e
d
c
on
t
ro
l
the
dot
patt
ern
s
of
the
ord
ered
dithe
r
pi
xe
l.
The
syst
em
adap
ts
t
wo
kilo
byte
s
of
hi
dd
e
n
data
for
a
bit
le
vel
(25
6*256)
im
a
ges,
c
om
pliant
a
volum
e
of
or
data
-
hid
i
ng
rati
o
of
one
da
ta
bit
to
f
our
cov
e
r
i
m
age
bits
[
1
7
]
.
I
n
the
sa
m
e
con
te
xt,
the
te
rm
ste
gan
og
raphy
fact
ually
incom
es
“co
ver
e
d
wri
ti
ng
”
as
stemm
ed
fr
om
Gr
eek.
It
can
be
def
i
ned
as
the
art
an
d
sci
ence
of
co
m
m
un
ic
at
ion
in
a
way
that
it
us
ed
t
o
hid
e
t
he
act
ual
it
y
of
com
m
un
ic
at
ion
.
A
ge
ne
ral
ste
ga
nogr
a
ph
y
syst
em
is
sh
ow
n
in
Fig
ure
1
a
s
gi
ven
by
[1
8
]
in the f
ollow
i
ng:
In
this
sect
i
on,
it
is
ex
pl
ai
ned
the
res
ults
of
resear
ch
a
nd
at
th
e
sam
e
tim
e
is
gi
ve
n
the
com
pr
e
hensi
ve
disc
us
si
on.
Re
s
ults
can
be
pr
ese
nte
d
in
fi
gures,
gr
aphs,
ta
bles
a
nd
ot
her
s
t
hat
m
ak
e
the r
ea
der u
nd
erstan
d
easi
ly
[
2,
5]
. T
he disc
us
sio
n
ca
n be
m
ade in
se
vera
l su
b
-
c
hap
te
rs.
-
Em
bed
ded <da
ta
-
ty
pe>: So
m
et
hin
g t
o be
h
i
dd
e
n
i
n
s
om
et
hin
g
el
s
e.
-
Stego
-
<
data
-
ty
pe>: T
he ou
t
put of the
hid
i
ng
process
, som
eth
in
g
that
has
th
e em
bed
ded.
-
Cov
e
r
<
data
-
t
ype>:
A
n
i
nput
with
“
or
igi
nal”
f
orm
of
the
ste
go
-
m
es
sage.
I
n
s
om
e
ap
plica
ti
on
s
su
c
h
a
cov
e
r
m
essage
is
gi
ven
from
the
ou
tsi
de
,
it
can
be
c
ho
s
en
durin
g
the
hid
i
ng
pro
cess.
T
he
le
tt
er
is represe
nted by the
das
he
d extensi
on to
t
he
inn
e
r hidi
ng
process
.
-
Stego
-
obj
ect
: t
he ou
t
pu
t
from
steg
o
-
syst
em
, s
om
et
hin
g hid
den in it.
-
Stego
-
key:
Ext
ra
secr
et
inf
or
m
at
ion
can
be
hidden
in
t
he
s
te
go
-
syst
em
th
e
sam
e
key
is
ty
pical
ly
req
uir
ed
to ex
t
ract t
he
e
m
bed
ded
m
essage a
gain.
-
The pr
ocess
of
hid
in
g
t
he
em
bed
de
d
m
essage
is cal
le
d
Em
bed
di
ng.
-
Gett
ing
t
he
em
bedde
d
m
essage out of t
he
ste
go
-
m
essage agai
n
i
s call
ed Ex
tract
ing
[19]
.
Figure
1.
Ge
ne
ral stega
nograp
hy syst
em
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3
5
5
0
-
3
5
5
7
3552
3.
PROP
OSE
D SYSTE
M
The
syst
em
a
tic
app
r
oac
h
is
us
e
d
inv
isi
ble
char
act
e
r
that
can
be
cl
assifi
ed
into
thr
ee
chief
sta
ges:
Hiding
(secr
et
)
te
xt
in
(cover
)
te
xt
us
i
ng
the
inv
isi
ble
c
har
a
ct
er
(
hid
in
g
ke
ys),
e
xtracti
ng
the
(
or
i
gin
al
)
t
ext
by
us
in
g
t
he hid
de
n keys a
nd p
e
rfor
m
ance ev
al
ua
ti
on
as
sho
wn in F
i
gure
2.
Figure
2
.
Flo
w
ch
a
rt of the
sy
stem
at
ic
ap
pro
ach
3.1.
Hidi
n
g te
xt in te
xt
Be
low
t
he
ste
ps
for
hid
i
ng
t
he
(secr
et
)
te
xt
in
(i
nnoce
nt)
te
xt.
C
hoos
e
any
te
xt
with
enou
gh
siz
e
as in
no
ce
nt tex
t t
o
use
it
as c
over
text.
T
he
h
i
de pr
oce
dures a
re as f
ollow
s:
-
Divid
e
the
(c
over
-
Te
xt)
to
b
l
ocks (n
um
ber
of cha
racters
or
byte
s).
-
In each
bl
ock hi
des on
e
ch
a
ra
ct
er of
the
sec
r
et
m
essage (
T
e
xt).
(Note:
both
of
them
co
ver
a
nd sec
ret m
essa
ge
c
har
act
e
rs
a
re r
e
presente
d i
n
A
SCII c
od
e
)
.
-
The
cal
c
ulati
on
s
of each
b
l
oc
k
le
ngth
is:
Bl
ock
le
ng
t
h
=
.
ℎ
−
.
ℎ
−
(Byt
e)
-
Fr
om
each
bloc
k
we
le
ave
tw
o
byte
s
(tw
o
f
r
om
the
beg
inn
i
ng
a
nd
at
the
e
nd
of
the
blo
c
k)
that
will
no
t
be
us
e
d
to
hi
de
the
secret
m
es
sage.
(
N
ote:
m
or
e
than
tw
o
byte
s
can
be
le
aved
to
in
crease
the
hi
di
ng
capaci
ty
).
Sear
ch
the
fi
rst
bl
ock
t
o
fin
d
a
n
identic
al
(A
S
CII)
value
t
o
that
of
t
he
first
secret
m
essage
char
act
e
r.
-
If
we f
ound it
then inse
rt in
vis
ible cha
racter t
o
the
r
i
gh
t
of it
to
in
dicat
e the
h
id
de
n
c
har
act
er.
-
Else
,
if
we
no
t
found
a
n
ide
ntica
l
(A
SC
II)
value
t
hen
w
e
search
agai
n
for
the
c
har
a
ct
er
that
is
ei
ther
gr
eat
er
t
han
s
ecret
char
act
e
r
(A
SC
II)
valu
e
by
(1
,
3,o
r
5)
or
le
ss
tha
n
secret
char
act
er
(ASCI
I)
va
lue
by (3
or 5); i
f we
fou
nd one t
hen inse
rt an i
nvisi
ble c
har
act
er to t
he rig
ht
of it
.
-
To
in
dicat
e
w
he
ther
th
e
cha
ra
ct
er
to
the
le
ft
of
i
nv
isi
ble
ch
aracte
r
was
ide
ntica
l
increase,
or
dec
rease
by
so
m
e
value.
The
fo
ll
ow
i
ng
prot
oco
l
(
dep
e
ndin
g
on
the
tw
o
byte
s
that
we
le
ft
at
the
beg
in
ning
an
d
e
nd
of the
blo
c
k
).
-
The
c
om
par
iso
n
bet
wee
n
the
cov
e
r
-
te
xt
an
d
ste
g
-
obj
ect
doesn’
t
fin
d
a
ny
m
od
ific
at
ion
(
befor
e
an
d
a
fte
r
hid
in
g
te
xt
)
a
nd
it
s
hould
be
ide
ntica
l
to
pr
e
ve
nt
any
dr
a
w
n
su
s
pici
on
of
the
at
ta
cker
s
e
xisted
into
innocent te
xt a
s sho
wn in Fi
gure
3.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Text
h
idi
ng in
text
u
sin
g
i
nvisible
char
acter
(
Nada A
bdul
Az
iz Must
afa
)
3553
Figure
3
.
Secre
t t
ext in t
he
to
p an
d
the
cove
r t
ext at t
he b
ottom
(
befor
e/
aft
er
hid
in
g
te
xt)
3.2.
Hidi
n
g
p
rocessin
g
The w
ho
le
ste
ps f
or
each
poin
t i
s explai
ned a
s foll
ow
s:
-
In
se
rt in
visible
ch
a
racter at t
he
end
of the
bl
ock to i
nd
ic
at
e
an
ide
ntica
l ch
aracte
r value
.
-
In
se
rt in
visible
ch
a
racter at t
he
b
e
ginnin
g of
the b
l
ock, m
eans t
hat the c
ha
r
act
er
val
ue
is i
ncr
ease
d by
one.
-
In
se
rt
in
visible
cha
racter
a
fte
r
the
fir
st
byte
of
the
blo
c
k
m
eans,
that
th
e
cha
racter
va
lue
is
i
ncr
ease
d
by th
ree.
-
In
se
rt
inv
isi
ble
char
act
er
afte
r
the
seco
nd
by
te
of
the
blo
c
k,
m
eans
that
t
he
cha
racter
va
lue
is
increas
ed
by
fi
ve.
-
In
se
rt
inv
isi
ble
char
act
er
be
f
ore
the
first
byte
fr
om
the
end
of
the
blo
c
k,
m
eans
that
the
char
act
e
r
value
is
decr
ease
d by t
hr
ee
.
-
In
se
rt
in
visible
char
act
e
r
be
f
or
e
t
he
byte
be
fore
the
sec
ond
byte
from
t
he
en
d
of
t
he
blo
c
k
m
eans
that
the ch
a
racter
va
lue is
decr
eas
ed by fi
ve.
Fo
r
eac
h
blo
c
k
re
peats
the
ste
ps
f
ro
m
(E
to
H)
unti
l
al
l
secret
char
a
c
te
rs
are
hi
dden
Fig
ur
e
4
,
The flo
w
c
har
t
for hidin
g
te
xt
in text as
sho
w
n
Fi
gure
5.
W
h
en
I:
is
the in
v
isib
le characte
r
an
d
C: is any
valu
e (
A
SCII)
.
Figu
re
4
.
Hi
ding text i
n
te
xt usi
ng in
visible c
har
act
er
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3
5
5
0
-
3
5
5
7
3554
Figure
5
.
Fl
ow
ch
a
rt to hide
te
xt in
te
xt
f
:
se
cret m
essage f
i
le
,
f
1: cover
f
il
e
,
b:
byte
,
A
: A
rr
ay
of c
ha
racter
BL
: B
l
ock Le
ng
t
h
Wh
il
e th
e
al
go
rithm
f
or
hid
i
ng
(secr
et
)
text i
n (in
no
ce
nt)
te
xt as
fo
ll
ows:
Sta
rt
Loo
p: w
hile (n
ot Eo
f
ste
go
-
ob
j
ect
f
il
e) d
o
Be
gin
Bl
ock
read
(
ste
go
-
ob
j
ect
f
il
e, b
, Block
leng
th)
Bl
ock
read
(
secret fil
e, c,
1)
Loo
p: Repeat i
=3;
I
f
C=
(
b [i]
or
(b [
i]
+1) o
r (b
[i]
+3)
or (b
[i]
+5 or
(b
[i]
-
3 o
r
(
b[i
]
-
5))
t
hen
Be
gin
b
[i+1] = In
visible character;
(b
[1
]
or
b[
2]
or
b[
3]
or
b[
blo
ck
le
ng
th]
or
b[
blo
ck
le
ng
th
-
1]or
b[
blo
ck
le
ng
th
-
2]
)
=
inv
isi
ble
char
act
er {accor
din
g
to ab
ov
e}
Fo
un
d
= tru
e;
End
;
Un
ti
l (f
ou
nd
)
3.3. Ex
tra
c
tin
s
t
ag
e
The
ext
racti
ng
sta
ge
is
inten
ded
t
o
rec
ov
e
r
the
hidde
n
te
xt
from
the
steg
o
-
obj
ect
.
E
xtr
act
ing
sta
ge
include
s
ext
rac
ti
ng
sec
ret
-
te
xt
from
the
ste
go
-
ob
j
ect
te
xt
tha
t
it
is
ind
ic
at
ed
as
the
op
posit
es
of
h
i
de
proc
edure
al
gorithm
s [
20]
. Th
e
w
ho
le
al
gorithm
is u
se
d
to
ex
t
ract t
he
secret te
xt
as s
how
n
in
d
et
ai
ls
as
bellows:
Start
Bl
ock
rea
d (ste
go
-
obj
ect
f
il
e,
b,
6o)
j= b [
1]
No
. O
f
par
t = b
[1
+j]
No
. O
f
By
te
in
each p
art = b
[1
+2j
]
an
d
b
[2
+2j
]
No
. O
f
byte
in
la
st par
t = b
[1
+3J]
Leng
th o
f
te
xt b
lock
= b
[
1+4
j
]
Loo
p: r
epeat
Loo
p: Fo
r
i=
1
to
(N
o.
o
f
byte
h
ide in each
p
art) N
S
Bl
ock
r
ead (
ste
go
-
ob
j
ect
, c,
1
)
If
(
i =A [
i]
)
then
Bl
ock
w
rite
(
extract fil
e, c,
1)
U
ntil
(
N
o.
Of pa
rts)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Text
h
idi
ng in
text
u
sin
g
i
nvisible
char
acter
(
Nada A
bdul
Az
iz Must
afa
)
3555
4.
E
X
PERI
MEN
TAL
RES
UL
T
Pr
efect
ste
ga
nogra
ph
y
is
w
hen
ge
ts
ste
go
-
obj
ect
sim
ilar
to
ori
gi
nal
innoce
nt
te
xt
by
bot
h
per
ce
ptu
al
ly
a
nd
com
pu
te
r
r
eadin
g.
This
m
ay
be
im
po
ssib
le
to
reac
h.
F
or
e
xp
e
rim
ental
res
ults,
it
wa
s
f
ound
that
the c
over
wh
e
n
it
c
hange
s to
ste
go
-
obj
e
ct
g
ives
the cl
ose
st crit
eria t
o
or
i
gin
al
c
ov
e
r.
Ther
e are two
d
iffer
ent sch
em
es that i
s u
sed
to g
ive an
idea if th
ese are a h
idd
en
m
essage o
r
no
t:
-
Bl
ind
sch
em
e
.
O
ne
is
the
process
t
hat
shou
l
d
be
possible
with
out
the
or
igi
na
l
cov
e
r
cal
le
d
(b
li
nd sc
hem
es)
w
hic
h need
a
high tech
ni
qu
e
to disco
ver an
d detec
t t
he hid
den m
essage [21
]
.
-
Escr
ow
schem
e.
Th
e
sec
ond
schem
e
wh
e
r
e
the
or
igi
nal
cov
e
r
si
gn
al
is
nee
ded
to
revea
l
the
hi
dd
e
n
inf
or
m
at
ion
call
ed
(esc
r
ow sc
hem
e).
The
sec
ond
sc
hem
e
co
m
par
e
betwee
n
th
e
ori
gin
c
over
an
d
ste
go
-
ob
j
ect
and
chee
k
if
there
a
re
a
ny
diff
e
re
nt.
The
seco
nd
sc
hem
e
is
us
ed
to
evaluate
the
perfor
m
ance
evaluati
on
by
ta
kin
g
s
ever
al
crit
eria
that
it
was use
d
to
eva
luate
the c
ove
rs
[
22]
.
4.1. A
voidi
ng
stego
-
anal
ys
is
The
dev
el
op
ed
al
go
rithm
fo
r
info
rm
at
ion
hid
ing
in
this
stud
y
was
im
plem
ented
ta
kin
g
into
accou
nt
ste
go
-
analy
sis.
This
al
go
rithm
includes
so
m
e
featur
es
that
m
ay
help
in
avo
iding
ste
go
-
analy
sis
at
ta
cks.
These f
eat
ur
es are:
-
Using
pri
vate
ste
go
-
ob
j
ect
to
pr
e
ven
t
dr
a
wi
ng
s
uspic
io
n
about
the
hi
dde
n
inf
orm
ation
.
A
com
par
iso
n
betwee
n
ste
go
-
obj
ect
an
d
the
inn
oce
nt
co
ve
r
m
ay
be
ineff
ic
ie
nt,
and
any
m
od
ific
at
ion
on
the
or
igi
nal
cov
e
r
m
ay
be
i
m
per
cepti
bl
e
[23].
Fi
gur
e
6
e
xp
la
in
com
par
ing
bet
ween
t
he
te
x
t
-
co
ver
an
d
it
s
corres
pondin
g
ste
go
-
te
xt.
Fro
m
the
res
ult
of
com
par
in
g
between
the
i
nnoc
ent
te
xt
a
nd
the
ste
go
-
ob
j
ect
,
it
o
bs
e
rv
e
d
t
ha
t anyo
ne
ca
nnot
see an
y m
od
if
ic
at
ion
in
steg
o
-
ob
j
ect
.
-
Using
m
ulti
pl
e
sta
ges
of
en
crypti
on
to
preven
t
the
extr
act
ion
of
the
hidden
in
f
or
m
at
ion
as
m
os
t
a
s
po
s
sible
[24], i
.e.
if
ther
e
is a
ny suspici
on a
bout the
h
i
dde
n
in
f
or
m
at
ion
, t
hey can
no
t
be
extracte
d.
Hidin
g
the
secret
te
xt
in
inn
ocen
t
te
xt
at
first
by
us
ing
the
m
et
ho
d
of
the
hyper
te
xt
m
ark
up
la
ng
uag
e
(H
TML
)
file
s.
The
fil
es
util
iz
ed
to
con
vey
the
info
rm
at
ion
m
eanw
hile
add
ing
sp
ace,
ta
bs
,
and
“i
nv
isi
ble”
char
act
er.
Fo
r
exam
ple,
hid
ing
an
Eng
li
sh
secret
te
xt
Figu
re
7
in
an
inn
ocen
t
te
xt
Figu
re
8
,
ste
go
-
ob
j
ect
with
hid
den
info
rm
at
ion
,
and
witho
ut
dr
awn
su
sp
ic
ion
by
us
ing
im
ag
e
(1
),
wh
il
e
Figu
re
9
exp
la
ins
the
evaluati
on
resu
lt
s w
it
h
hid
den
inf
or
m
at
ion
w
hen
u
sing
stego
-
ob
j
ect
.
Fr
om
the
exp
erim
ental
resu
lt
s,
it
can
cl
early
ob
serv
ed
that
the
com
par
ison
resu
lt
of
per
fo
rm
ance
evaluati
on
(av
erag
e,
var
ia
nt,
ener
gy,
entro
py,
sim
il
ari
ty
,
secrecy
and
SN
R)
between
the
cov
er
-
te
xt
and
the
ste
go
-
ob
j
ect
is
a
li
tt
le
bit
diff
eren
t
that
it
ind
ic
at
e
anyon
e
can
no
t
seen
the
hid
den
op
erati
on
s.
The
aver
age
value
of
the
pr
op
os
ed
te
chn
iqu
e
carried
ou
t
bette
r
per
fo
rm
ance
with
an
increase
of
15
%%
com
par
ed
to
the av
erag
e v
al
ue
of
o
f
ener
gy an
d
sensiti
vity
[2
5]
us
ing
the BMP im
age stegan
og
rap
hy
.
Figure
6
.
Com
par
is
on b
et
wee
n
the
origi
n
(te
xt
-
c
ov
e
r) an
d
i
ts corres
pondin
g
ste
go
-
ob
j
ect
te
xt
ORIGIN
TE
XT
STE
GO TE
X
T
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3
5
5
0
-
3
5
5
7
355
6
Figure
7. En
glish te
xt (
sec
ret
m
essage)
,
whi
ch wil
l
be hid
den in te
xt
Figure
8. The
innoce
nt text c
over
, in w
hich
inf
or
m
at
ion
w
i
ll
b
e h
i
dd
e
n
Figure
9. Com
par
is
on b
et
wee
n
te
xt
-
co
ver a
nd ste
go
-
ob
j
ect
te
xt d
e
pends
on th
e
ev
al
uatio
n
m
et
rics
5.
CONCL
US
I
O
N
Hiding
te
xt
in
te
xt
us
in
g
in
visible
char
act
e
r
pro
vid
es
a
n
effi
ci
ent
m
et
ho
d.
This
stu
dy
was
perform
ed
in
th
ree
sta
ges
,
hi
de
te
xt
in
t
ext,
e
xtract
th
e
secret
te
xt
f
r
om
the
ste
go
-
obj
ect
te
xt
an
d
e
valuati
on
m
et
rics.
The
hid
e
proc
edure
ca
n
pe
rfor
m
a
secur
it
y
o
pe
rati
on
in
order
t
o
hi
de
the
secret
te
xt
.
The
e
xtract
sta
ge
al
gorithm
rev
erses
the
hi
de
proce
dure
ef
fe
ct
s
in
order
to
extract
the
hi
dd
e
n
te
xt.
Mo
reover
,
us
i
ng
public
ste
go
-
ob
j
ect
to
pr
e
ve
nt
knowing
the
or
i
gin
al
co
ver,
in
orde
r
to
preve
nt
a
ny
com
par
ison
be
twee
n
the
in
no
ce
nt
cov
e
r
an
d
ste
go
-
obj
ect
.
T
he
re
are
m
any
l
evels
of
com
plexity
with
the
propose
d
sy
stem
,
wh
ic
h
preve
nt
at
ta
cker
to reac
h
the
secret m
e
ssage. C
onse
quently
, th
e
cap
abili
ty
to
incre
ase the ca
pacit
y of
hid
i
ng. Pr
opos
e
d
syst
e
m
dep
en
ds
on
m
od
ify
i
ng
(
2
-
3)
bit
of
s
ecret
m
essage.
Also
ca
n
i
ns
te
ad
of
represe
nt
each
c
har
act
er
wit
h
(8
-
bit)
can
r
e
present it
w
it
h (
5
-
bit)
(
no. of
c
ha
racter a
re
26) a
nd h
e
nce
(
3
-
bi
t) sav
e ca
n ac
c
om
plished
.
REFERE
NCE
S
[1]
V.L
.
Nar
a
y
ana,
A.P.
Gopi,
an
d
N.
A.
Kum
ar
,
"D
iffe
ren
t
t
e
ch
nique
s
for
hidi
n
g
the
te
x
t
infor
m
at
ion
using
tex
t
stega
nogra
ph
y
t
e
chni
ques:
A surv
e
y
,
"
Ing
eni
eri
e
d
es
Syste
mes
d'Inf
orm
ati
on
,
vol
.
2
3,
pp
.
115
,
2018
.
[2]
A.F.M.A.
K.
Sana
and
S.
Mohmm
ad,
"A
n
F
PG
A
implementa
ti
on
of
sec
ure
d
steganograph
y
communic
a
ti
on
s
y
st
em,"
Tikrit
Journal
of
Engi
n
ee
ring
Scienc
es
,
vol
.
19
,
p
p.
14
-
23
,
2012
.
[3]
A.
Kum
ar
and
K.
Pooja
,
"S
te
gan
ogra
ph
y
-
A
data
hidi
ng
t
ec
hniqu
e
,
"
Inte
rnat
ional
Journal
of
Comp
ute
r
Application
s,
vol.
9
,
pp
.
19
-
23
,
2010
.
[4]
H.
Mohaje
r
i
M
oghadda
m
,
B
.
Li
,
M.
Der
akhsha
ni,
and
I.
Go
ldbe
rg,
"S
k
y
p
e
m
orph:
Protocol
obfusca
t
ion
for
tor
bridge
s,"
in
Pro
ce
ed
ings o
f
the 2
012
ACM conf
ere
nce on
Comput
er
and
communi
cat
ions se
curit
y
,
pp.
97
-
108
,
201
2.
[5]
I.
Sede
eq, "HT
ML
Stega
nogr
ap
h
y
Algorit
hm
s a
nd
Detect
ion
Me
thods,"
Univer
si
t
y
of
L
ive
rpoo
l,
2
018.
[6]
S.M.A.
Al
-
Nofai
e
and
A.
A.
-
A.
Gutub,
"U
tiliz
ing
ps
eudo
-
spac
es
to
improve
Arabi
c
te
x
t
stega
nogra
ph
y
for
m
ult
imedia
d
ata com
m
unic
at
ions
,
"
Multimedia
Tools
and
App
li
ca
ti
ons,
pp
.
1
-
49
,
2
019.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Text
h
idi
ng in
text
u
sin
g
i
nvisible
char
acter
(
Nada A
bdul
Az
iz Must
afa
)
3557
[7]
G.
J.
Sim
m
ons,
"The
pr
isoners’
p
roble
m
and
the subli
m
ina
l
cha
nn
el
,
"
in
Adv
an
ce
s
in
Cryptology
,
p
p.
51
-
67
,
1984
.
[8]
T.
Pevný
,
T
.
Fill
er
,
and
P.
Bas,
"U
sing
hi
gh
-
dimensional
image
m
odel
s
to
per
form
hi
ghl
y
undetec
ta
b
le
stega
nogra
ph
y
,
"
in
Int
ernati
onal
Workshop on
Inf
orm
ati
on
Hiding
,
pp
.
161
-
177
,
2
010.
[9]
M.
Kharra
z
i,
H
.
T.
Senca
r
,
and
N.
Mem
on,
"Im
age
stega
nogr
aph
y
and
st
eganal
y
s
is:
Conc
ep
ts
and
pra
ct
i
ce,
"
in
Math
emati
cs
and
Computat
ion
in
Imaging
Scienc
e
and
Information
Pr
oce
ss
ing
,
Ed
:
World
Scientifi
c
,
pp.
177
-
207
,
20
07.
[10]
M.
Mishra,
G.
Ti
wari
,
and
A.
K.
Yada
v,
"S
ec
re
t
comm
unic
at
ion
using
publi
c
ke
y
stega
nogra
ph
y
,
"
in
Inte
rnation
al
Confe
renc
e
on
R
ec
en
t
Ad
vances
and
Innov
at
ion
s
in
Eng
ine
ering
(
ICRA
IE
-
2014)
,
pp.
1
-
5
,
2014
.
[11]
G.
Sw
ai
n
and
S.
K.
La
nk
a,
"A
quic
k
rev
i
ew
of
net
work
sec
ur
ity
and
steg
ano
gra
ph
y
,
"
Inte
rna
ti
onal
Journal
o
f
El
e
ct
ronics
and
Computer
Scien
ce
Engi
ne
ering
,
vol.
1
,
pp
.
426
-
4
35,
2012
.
[12]
M.
Douglas,
K.
Baile
y
,
M.
L
ee
ne
y
,
and
K.
Curra
n,
"A
n
ov
erv
ie
w
of
steg
a
nogra
ph
y
techni
ques
appl
i
ed
to
the
pro
te
c
ti
on
of
biometric
d
at
a
,
"
Multimedia
Too
ls and
App
li
ca
tions
,
vol. 77, pp.
17333
-
17373,
2
018.
[13]
R.
Böhm
e,
"P
rinc
iples
of
m
oder
n
steg
anogr
ap
h
y
and
st
ega
n
aly
sis,"
in
Ad
van
ce
d
S
tat
ist
ic
al
Ste
ganal
ysis
,
Ed
:
Springer
,
pp
.
11
-
77,
2010
.
[14]
S.
Ta
n
and
B.
L
i,
"T
arg
eted
steg
ana
l
y
sis
of
edge
ada
pt
ive
image
stega
nogra
ph
y
base
d
on
LSB
m
at
chi
ng
r
evi
sit
ed
using B
-
spline fi
tt
ing,
"
IE
EE
S
ig
nal
Proc
essing
Letters
,
vol
.
19
,
p
p.
336
-
339
,
201
2.
[15]
J.
Zha
ng,
H.
Hu
ang,
L
.
W
ang,
H
.
Li
n
,
and
D.
Ga
o,
"Coverl
ess
Text
Inform
at
ion
Hiding
Method
Us
ing
the
Freque
nt
W
ords Hash,
"
I
nte
rnational
J
our
nal
Ne
twork
Se
c
urity
,
vol
.
19
,
pp
.
1016
-
1023
,
20
17.
[16]
M.
S.
Hana
'a,
"
A
Natur
al
La
ng
uage
St
ega
nogr
aph
y
T
ec
hniqu
e
for
T
ext
Hid
in
g
Us
ing
LSB'
s,
"
Engi
ne
ering
a
nd
Technol
ogy
Jour
nal
,
vo
l. 26, pp.
351
-
364,
2008
.
[17]
R.
Poornim
a
and
R.
Isw
ar
y
a,
"
An
over
vie
w
of
digi
t
a
l
image
s
te
ganogr
aph
y
,
"
I
nte
rnational
Jou
rnal
of
Computer
Sci
en
ce and
Eng
ine
ering
Surve
y
,
vol.
4
,
p
p
.
23
,
20
13.
[18]
L.
Y.
Por
and
B
.
Deli
n
a,
"Inform
at
ion
h
idi
ng:
A
new
appr
oa
ch
in
te
xt
steg
anogr
a
ph
y
,
"
in
7
th
WS
EA
S
In
t.
Con
f.
o
n
Appl
ie
d
Comput
er
&
Appl
ie
d
Computati
onal
Scie
nce
(
Ac
a
cos
'08)
,
Hang
zhou, China,
2008
[19]
G.
Nehru
and
P.
Dhar,
"A
det
ai
l
ed
look
of
audi
o
stega
nogra
ph
y
technique
s
using
LSB
and
gene
ti
c
a
lgori
t
hm
appr
oac
h
,
"
Int
er
nati
onal Journal
of
Comput
er
Sc
i
enc
e
Iss
ues
(
IJCSI
)
,
vol.
9
,
p
p
.
4
0
2,
2012
.
[20]
C.
Ba
ldwin,
A.
Mac
Corm
ac
k
,
and
J.
Rusna
k,
"H
idde
n
str
uct
ure
:
Us
ing
net
work
m
et
ho
ds
to
m
ap
s
y
s
te
m
arc
hi
te
c
ture,"
R
e
search
Policy
,
v
ol.
43
,
no
.
8
,
pp
.
1381
-
1397,
201
4.
[21]
G.
C.
Kess
le
r
a
nd
C.
Hos
m
er,
"A
n
over
vie
w
of
stega
nog
ra
ph
y
,
"
in
Adv
an
ce
s
in
Comput
ers
,
Ed
:
El
se
vier
,
vol.
83
,
pp
.
51
-
1
07
,
2011
.
[22]
A.A.J.
Alt
aay
,
S.
B.
Sah
ib,
and
M.
Za
m
ani,
"A
n
int
roduc
ti
on
to
image
steg
anogr
aph
y
techniqu
es,
"
in
Inte
rnat
ion
al
Confe
renc
e
on
A
dvanc
ed
Comput
er
Scienc
e
App
licati
ons and
Tec
hnologi
es
(
ACS
AT)
,
pp.
122
-
12
6
,
2012
.
[23]
C.
Sum
at
hi,
T
.
Santa
nam,
and
G.
Um
amahe
s
wari
,
"A
stud
y
of
var
ious
ste
ganogr
aphic
tec
hnique
s
used
for
informati
on
h
iding,"
arXi
v
prepr
int
arXi
v:1
401.
5
561,
2014
.
[24]
A.
Bir
y
ukov
,
I.
Pus
toga
rov,
an
d
R.
-
P.
W
ei
nm
ann,
"Tra
wl
ing
for
tor
hidde
n
services:
Detect
i
on,
m
ea
surem
ent,
dea
non
y
m
iz
a
ti
o
n,
" i
n
2013
I
EEE
Symposium on
Sec
uri
ty and
Pri
vac
y
,
pp
.
80
-
94
,
2013
.
[25]
S.
Alsale
em,
"A
utomate
d
Arabi
c
Te
x
t
Cat
eg
oriz
a
ti
on
Us
ing
SV
M
and
NB,
"
Int.
Arab
J.
e
-
Technol.,
vol
.
2
,
pp.
124
-
128
,
20
11.
BIOGR
AP
H
Y
O
F
AU
TH
OR
Nad
a
Abdu
l
A
z
i
z
Mustafa
,
her
born
at
Septem
ber
1
st
1966,
Scie
nti
f
ic
D
egr
e
e:
Ms
c.
i
n
computer
sci
ence.
R
ese
ar
ch
Proj
ec
ts
in
Th
e
Fel
i
d
of
Spec
ia
l
iz
a
tion
to
Th
e
Envi
r
onm
ent
and
Socie
t
y
or
the
Deve
lopment
of
Educ
a
ti
on:
(
a)
Design
and
Im
ple
m
ent
at
io
n:
Propos
ed
Enc
oding
and
Hiding
Te
xts
in
an
Im
age
,
Uni
ver
sit
y
of
Sul
y
m
ai
na,
2010.
(b
)
Util
izin
g
Com
pute
rs
in
Te
a
chi
ng
,
Univ
ersity
of
B
agh
dad,
Conf
ere
nc
e
,
2010
.
(c)
A
proposed
Te
chn
ique
for
Inform
at
ion
Hiding
B
ase
d
on
DCT,
IJ
ACT,
South
Korea
,
2011.
(d)
Enc
r
y
p
ti
on
a
te
x
t
using
aff
ine
c
iphe
r
and
hidi
ng
i
t
in
the
col
ore
d
image
b
y
using
the
Quant
iz
a
ti
o
n
stage
,
Univ
er
sit
y
of
Mus
ta
ns
iriy
a
,
Confer
ence
,
2013
.
(
e)
Fac
e
detec
t
ion
and
th
e
eff
e
ct
of
cont
rast
and
bright
ness
,
JNIT
,
South
Korea
,
2014.
(f)
The
Eff
ect
of
the
Sm
oothi
ng
Filt
er
on
an
Im
age
Enc
r
y
pt
ed
By
the
Blowfish
Algorit
hm
The
n
Hiding
It
in
A BMP
Im
age
,
J
NIT
,
South
Kore
a
,
2014
.
Evaluation Warning : The document was created with Spire.PDF for Python.