TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.5, May 2014, pp
. 3323 ~ 33
3
0
DOI: http://dx.doi.org/10.11591/telkomni
ka.v12i5.4929
3323
Re
cei
v
ed O
c
t
ober 1
8
, 201
3; Revi
se
d Novem
b
e
r
28, 2013; Accept
ed De
cem
b
e
r
18, 2013
Application of Remote Sensing Axis Line Method in
Xiaomiaohong Creek
Yan Gu*
1
, Ming-jie Gong
2
, Huan Li
3
1
Colle
ge of Ge
ogra
phic a
nd B
i
olo
g
ic
al Inform
ation,
Na
nji
ng
Univers
i
t
y
of Post and T
e
leco
mmunicati
on,
Nanj
in
g 21
002
6, Chin
a
2
Jiangs
u Acad
em
y
of En
gin
e
e
rin
g
of Surve
y
ing a
nd Ma
ppi
ng, Nan
jin
g 21
001
3, Chi
n
a
3
Colle
ge of Ge
ogra
phic Sci
e
n
c
e, Nanj
ing N
o
rmal Univ
ersit
y
, Nanji
ng 2
100
26, Chi
n
a
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: gu
yan
@
nj
upt
.edu.cn
A
b
st
r
a
ct
T
he traditi
on
a
l
researc
hes o
n
evol
utio
n of ti
dal creek
h
a
ve be
en l
i
m
it
ed for the rea
s
on that
precis
e meas
u
r
ed d
a
ta be r
e
q
u
ire
d
. On the o
t
her ha
nd,
the
Re
mote S
ensi
ng Axis L
i
n
e
method (
R
SAL)
can
mo
nitor the ev
oluti
on of the tidal cre
e
k acc
u
ratly an
d pro
m
ptly. T
h
is pa
per extr
acted t
he Axis lin
es of
Xiao
mi
ao
ho
ng
creek
base
d
o
n
o
n
e
MSS i
m
age
a
nd s
i
x
T
M
i
m
ag
es
in
the
peri
o
d
of 1
979, 19
87, 19
89,
199
3, 199
5, 20
03 a
nd 2
0
0
9
w
i
th RSAL a
nd t
ook a furt
h
e
r a
nalysis. T
he r
e
sults show
ed t
hat: On the ba
sis
of over
layi
ng
g
ener
ated
axis
l
i
ne
an
d
DEM (
d
igit
al
elev
atio
n
mo
del)
d
a
ta
both
acq
u
ire
d
i
n
2
003,
axis
li
n
e
can be the su
b
s
titution of the thal
w
eg li
ne be
cause of si
mil
a
r trend and
min
o
r devi
a
tion of
abov
e tw
o kinds
of lines, w
h
ich
is reveal
ed th
at RSAL w
a
s service
abl
e an
d effective to
und
erstan
d dy
na
mic ev
oluti
o
n of
tidal cr
eeks
qui
ckly an
d acc
u
r
a
tely. T
he
upp
er sectio
n of Xi
ao
mi
aoh
on
g cr
eek w
a
s d
epos
iting
an
d grow
i
n
g
all the ti
me a
nd sub
o
rd
inat
e brach
e
s w
e
re grad
ua
lly d
i
sap
pear
ed, th
e south si
de
of mi
ddl
e sect
ion
possess
ed rel
a
tively strong
er hydro
d
yna
m
ic force an
d much
more su
bor
din
a
te branc
hes w
e
re dev
elo
p
i
n
g
.
Ke
y
w
ords
:
remote s
ensi
ng a
x
is line
method
, tidal creek, Xi
ao
mi
aoh
on
g creek, Lvsi port
Copy
right
©
2014 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
The
radial
sand
ridg
es o
f
South Yell
ow Se
a lie
s
in the mi
ddl
e coa
s
tal of
Jian
gsu,
betwe
en Xinyangg
ang an
d
Yaowan
gga
n
g
, which is
be
tween a
bolish
ed Yellow
River und
erwate
r
delta an
d Ya
ngtze
Rive
r u
nderwate
r d
e
l
ta. The radi
a
l
san
d
ridge
s stret
c
h f
r
om
so
uth to n
o
rth,
200
km lon
g
from so
uth to
north, 90
km
wide from
ea
st to we
st, with Jian
gga
ng
as its
start p
o
i
nt.
The wh
ole se
a area i
s
abo
ut 2×104
km
2
.
About 70 sa
nd ridg
es con
s
ist of the wh
ole san
d
ridg
es
grou
p, 10 of
whi
c
h a
r
e a
b
o
ve the o
c
ea
n su
rfa
c
e
in l
o
w tide, exp
a
nding
clo
c
kwi
s
e from no
rth
to
south li
ke a sector [1, 2].
It is a prereq
uisite to
reali
z
e the i
m
po
rtanc
e of chan
ge of subm
arine reli
ef, whi
c
h
ca
n
help to expl
oit the resou
r
ce
s al
ong t
he co
astlin
e and mud flat
, protect the
eco
s
ystem
and
environ
ment i
n
tidal flat a
n
d
wet land,
a
s
well
a
s
sci
entific-effe
ctively mana
ge t
he o
c
ea
n. Th
e
desi
r
e of Jia
ngsu Provin
ce Govern
me
nt is urge
nt to build the p
o
rts an
d ind
u
stry pa
rks a
nd
exploit the re
sou
r
ces
of th
e mud flat. T
he ra
dial
san
d
ridg
es i
n
S
outh Yello
w
Sea are the
key
zon
e
in
both
resea
r
ch an
d
develop
ment
. There
i
s
a
necessity to
unde
rsta
nd t
he featu
r
e
s
a
n
d
variation l
a
w of the rad
i
al sa
nd
rid
ges’ te
rrain
and l
and
scap
e scie
ntifically by u
s
ing
corre
s
p
ondin
g
techniq
u
e
s
, in order to p
r
ovide the su
pport in ba
sic material and
techniqu
e for
a
good pl
an in reasona
ble de
velopment.
Lvsi po
rt i
s
one of th
e radial
san
d
ri
dge p
o
rt
s, h
a
s
gre
a
t pot
ention in
dev
elopme
n
t.
There is a
ne
ce
ssity to un
derstand th
e
feature
s
an
d variation
la
w of
the
terrai
n
and
lan
d
sca
p
e
arou
nd the
L
v
si po
rt scie
ntifically by u
s
ing
co
rresp
ondin
g
techni
que
s, in o
r
de
r to provide t
h
e
sup
port in b
a
s
ic m
a
terial
a
nd tech
niqu
e
for a go
od pla
n
in rea
s
o
nab
le develop
me
nt. As the mai
n
cha
nnel affe
cting Lvsi p
o
rt, Xiaomia
ohon
g ch
ann
el has
relati
vely indepen
dent wate
r-sand
system. T
he revolution of X
i
aomiao
hon
g
cha
nnel
w
ill i
m
pact
on the
con
s
tructio
n
of the Lvsi
po
rt
dire
ctly.
The tradition
al re
se
arche
s
on evlutio
n
of
tidal cree
k have b
een li
mited for th
e
rea
s
o
n
that preci
s
e
measured da
ta be requi
re
d. On t
he other ha
nd, the Remote S
ensi
ng Axis Lin
e
method
(RS
A
L) can mo
n
i
tor the evolu
t
ion of t
he tidal cree
k a
c
curatly and p
r
omptly [3]. This
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3323 – 33
30
3324
pape
r took a f
u
rthe
r analy
s
i
s
of RSAL an
d applie
d to Xiaomiao
hon
g cha
nnel.
2. Stud
y
Are
a
and Da
ta
Acquisition
2.1. Stud
y
Area
Xiaomiao
hon
g chan
nel a
s
a rep
r
e
s
en
tative tidal cree
k of radia
t
e sand ri
dg
es area
usu
a
lly go
es
throug
h
sub
m
erg
e
n
c
e, a
nd that
pa
rall
els
with
co
ast
line a
nd
orie
n
t
ates in
a
NWW-
SEE direction. The lengt
h of this channel is
about
38km and width of
its entrance and mid-
segm
ent i
s
a
pproxim
ate 1
5
km
and
4.5
k
m respe
c
tively, its tail disap
pea
rs i
n
Rud
ong
sh
o
a
l.
Several subm
arine
sho
a
ls
are devel
ope
d in this are
a
[4].
Figure 1. Re
search Are
a
of Xiaomiaoh
on
g Cre
e
k
2.2. Data
Ac
quisition an
d Preproce
s
s
ing
In order to
u
nderstan
d th
e mig
r
ation
and
ch
ang
e of
Xiaomia
oh
ong cree
k
i
n
histo
r
y
perio
d, 1
sce
ne of MSS
(Multi Spect
r
al
Scan
ner
) im
agery
ca
ptured in
197
9 a
nd 6
scen
es
of TM
(Them
atic
M
appe
r)
im
ag
erie
s captu
r
ed
bet
wee
n
1987 and 2009 we
re colle
cted,
a
b
o
ve
image
rie
s
are
all over a rel
a
tively low tidal con
d
ition
s
with clo
ud fre
e
. Image sp
e
c
ific pa
ram
e
ters
see Ta
ble 1.
Remotely se
nse
d
ima
geri
e
s we
re geo
metric
ally pro
j
ected
to Ga
uss Kruge
r p
r
oje
c
tio
n
(GK-21 zone
) and
geo
re
ctified to WG
S-84
coo
r
din
a
tes system based
on 23
grou
nd
cont
rol
points (G
CPs),
whi
c
h are easily re
cogn
ized and st
ab
le ground
fea
t
ures an
d me
asu
r
ed
via RTK
GPS in
study
are
a
. A q
u
a
d
ratic polyn
o
m
ial fun
c
tion
and
nea
re
st
neigh
bor resampling
meth
od
were e
m
ploy
ed to
re
ctify RS ima
g
e
s
a
nd g
uarant
ee
the
root-mea
n-squa
re
(RMS) le
ss tha
n
0.
5
pixel. In this
process, the
sa
me bat
ch
of GPCs were
utilized i
n
each
RS image,
whi
c
h
means if
rectified
e
rro
r occu
rred a
n
d
led
to the
misregi
st
ratio
n
,
the same
error wo
uld be
p
r
od
uced
to
every ima
gery, to some
e
x
tent this
wa
y can
avoid t
he vari
ant e
r
ror p
r
o
bably
caused
by usi
n
g
different batch of GCPs, while ca
rryin
g out
the history trend analy
s
is of tidal cre
e
ks.
Table 1. Sele
cted Image
s
of Xiaomiaoh
ong Cree
k
Images
Spatial resolution
Capture time
MSS 57m
1979.8.4
Landsat TM
30m
1987.5.18
Landsat TM
30m
1989.8.11
Landsat TM
30m
1993.3.31
Landsat TM
30m
1995.4.6
Landsat TM
30m
2003.3.31
Landsat TM
30m
2009.4.28
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Applicatio
n of Rem
o
te Sensing Axi
s
Lin
e
Method in
Xiaom
iaohon
g Cre
e
k
(Yan
Gu)
3325
3. The Conc
ept of
Axile-l
ine Method
Becau
s
e
of the bathymetry is difficult to me
a
s
ure, so the migrati
on of tidal creeks is
hardly dete
r
mined a
s
we
ll. Here, axile line co
nce
p
t was introd
uce
d
in our
rese
arch an
d
the
c
h
arac
teris
t
ic of this
method is
to discuss
t
he
cha
n
ge of groun
d
feature
s
in t
w
o dim
e
n
s
io
nal
spa
c
e
s
, be
ca
use axile
-line
as a typical feature
c
an
re
pre
s
ent the shape of tidal
cre
e
k. If we can
extract many
axile-line
s
of our targ
et tidal cre
e
k in hi
stori
c
pe
riod,
the migratio
n and chan
g
e
of
this cree
k sh
ould be e
a
sily
monitore
d [5,6].
3.1. Extraction of Axile-line
In hydra
u
lic scien
c
e, th
e th
alwe
g of
river is
alway
s
co
ndu
cted to
an
alyze th
e evol
ution of
river
cha
nnel
and
pre
d
ict
its pote
n
tial
moving
tren
d in future.
Thalweg i
s
defined
as the
con
n
e
c
ting line of those
points, which
locate in
th
e cro
s
s se
cti
ons of rive
r and with d
e
e
pest
water
depth.
From the d
e
finition of thalweg we
can se
e,
acqui
sition of
thalweg se
verely
depe
n
d
s
upon
water d
epth data, as well kn
own to all water
d
epth is ha
rdl
y
extracted from RS imag
e,
whi
c
h m
ean
s it is imp
o
ssible to a
c
q
u
i
r
e thal
we
g o
f
tidal cree
k only relyin
g
on
RS ima
ge.
Ho
wever, it does n
o
t mea
n
we have n
o
cha
n
ce
to
monitor the chang
e of tidal cree
ks by RS
image, be
cau
s
e the ba
si
c idea of thalweg gives u
s
an inspiratio
n
,
so a simila
r sub
s
titute was
come
out a
n
d
introd
uced fo
r un
derstan
di
ng the
evolut
ion of tidal
cre
e
ks, whi
c
h
is
axile line
wha
t
we have proposed before and it will be us
ed in tidal
creek evol
ution research.
Re
cently, two kin
d
s of
way are
widely
us
e
d
for ext
r
actin
g
axile-l
ine of polygo
n
[7, 8],
mean
while, l
o
ts of
softwa
r
e, such a
s
R2V a
nd A
r
cGIS have b
e
en devel
ope
d
and
are abl
e to
gene
rate axil
e-line
rapidly
and a
c
curatel
y
[9]. An
exte
nsio
n modul
e
Arcsca
n in ArcGIS packag
e
wa
s an alternative way to extract axile-line of
Xiaomiao
hon
g tidal cre
e
k i
n
terms of its
compatibility with rem
o
te sensi
ng data.
3.2. The Process o
f
Extr
acting
Axile-lines
(1)
Wate
r bod
y separation
Tidal cree
k should be
se
parate
d
from
raw
imag
e before extracting axile-line
of out
targets.
Re
ce
ntly, many re
sea
r
ch p
r
op
o
s
ed
a lo
t
of ef
fec
t
ive methods
for
s
e
parating water
body,
among th
em
threshold i
s
a mostly em
ployed on
e. The water b
o
d
y sepa
ratio
n
prog
re
ss
are
as
follows: takin
g
a TM imag
e captu
r
ed i
n
1995 a
s
a
n
example, several a
r
bitra
r
y profile
s were
extracted
ove
r
TM5, a wav
eban
d cente
r
ed
at 1650
n
m
. We fo
und
that the val
u
e of
DN (Digi
t
al
Numb
er)
had
an
obviou
s
variation
ne
arb
y
the
wate
r li
ne, e
s
pe
cially
from
on
e
sid
e
cro
s
s a
noth
e
r
side,
and
fin
a
lly a th
re
sh
old valu
e 2
5
wa
s dete
r
mi
ned to
b
u
ild
mask file
for the
purpo
se
of
sep
a
ratin
g
water body.
(2) Axile-li
ne
extraction
Firstly, ma
sk
file sho
u
ld
be
pre
pared
already
for wate
r bo
dy sepa
ra
tion (Fig
ure 2
)
, whi
c
h
is bina
ry ima
ge only co
nta
i
ning bla
c
k a
nd white
colo
r and corre
s
p
ondin
g
value is set to 0 an
d 1.
Secon
d
ly, input the resam
p
led bina
ry image into
ArcGIS and loa
d
Arcscan m
odule, mea
n
w
hile
cre
a
te an
d l
oad a
new fi
le whi
c
h a
r
e
sha
pe form
at, keep
edit
con
d
ition an
d active Arcscan
module. Sel
e
ct vecto
r
iza
t
ion module
and set
co
ndition
s befo
r
e ru
nnin
g
it. And then two
para
m
eters
should b
e
set,
whi
c
h a
r
e “m
aximum li
ne width” and
“fo
r
mer ba
ckgro
und
color”.
After
finishin
g tho
s
e setting, axile-line can
be extr
acte
d automati
c
a
lly.Because
of an autom
atic
extraction
me
thod we utilized, so l
o
ts of
redu
nda
nt
de
tails are al
so
gene
rated
co
ncu
r
rently, su
ch
as tiny
axile-li
nes in
sub
o
rd
inate b
r
an
ch
e
s
of tid
a
l
cre
e
ks,
so
it
is
ne
ce
ss
ary
t
o
de
let
e
t
h
o
s
e lin
es
for sub
s
e
que
nt analysi
s
.
a)
b)
c)
Figure 2. Extraction Process of Axis Line
in
Xiaomiaoh
ong Cree
k (1
995) a
)
DN hi
stogram, b) t
he
binary ima
ge,
c) the axis lin
e
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3323 – 33
30
3326
3.3. Anal
y
s
is
of Axile-line
In orde
r to analyze the
migratio
n an
d cha
nge of
tidal cre
e
k
of Xiaomiaoh
ong, the
rep
r
e
s
entativ
ene
ss
of axile-line
s
sho
u
l
d
be eval
uat
ed firstly. The criterio
n of
the judgm
en
t is
wheth
e
r
axile
-line
ca
n m
a
tch th
alweg v
e
ry
well.
Th
e
way
wh
at we have
u
s
ed
is to
overl
ap
the
gene
rated axi
l
e-line d
e
rive
d from TM image (2
003
) o
v
er topograp
hic map me
a
s
ured in 200
3
,
so
it is rational t
o
comp
are the differen
c
e b
e
twee
n ea
ch
other in the
same year. Th
e re
sults (Fig
ure
3) sho
w
ed th
at the similar trend could
be f
ound b
e
twee
n extra
c
ted axile-lin
e
and me
asure
d
thalwe
g. In orde
r to ana
lyze the rela
tionshi
p between axile-li
n
e and thalweg cle
a
rly, three
rep
r
e
s
entativ
e p
r
ofile
s
were dete
r
min
e
d
.
Figure 3
(
a
)
sho
w
e
d
that
profile
one
crossed
axile-li
n
e
s
and th
e inte
rse
c
tion
s of
profile 1
and
ge
nerate
d
axile
-line
were n
a
m
ed to
A a
n
d
B from
no
rth
to
south. P
r
ofile 2 a
nd 3
cro
s
sed
axil
e-line
a
s
we
ll and i
n
tera
ction
s
of the
m
are
C an
d D
r
e
spec
tively.
(a)
(b)
(c
)
(d)
Figure 3. anal
ysis of axis li
nes (a
) secti
ons’ lo
ca
tio
n
(b)
se
ction 1 (c) se
ction 2
(d)
se
ction 3
From
Figu
re 3(b
)
we ca
n see
that
bl
ue
cu
rve
i
s
the
elevation
of p
r
ofile 1
from
south
to
north di
re
ctio
n, two re
d p
o
ints a
r
e inte
ract
io
ns
between thal
weg
and axile
-line
labeled
with
A
and B, ideally
point A and
B shoul
d be l
o
cate
d at
the
bottom of U-bend a
nd its
elevation sho
u
ld
be relatively lowe
r tha
n
th
ose
point
s lo
cated
at two
side
s, at thi
s
con
d
ition the
axile-line
can
be
treated a
s
th
alwe
g and b
e
able to re
p
r
esent the sh
ape of tidal cree
ks. In real
ity the axile-line
almost
can
n
o
t
match thalweg accu
rately
as the re
sult
of the intersecting
su
rface of cre
e
k i
s
not
symmetri
c
, in
anoth
e
r
wo
rds it
doe
s
no
t like ‘
U
’,
so
t
he bi
as is i
n
evitable. Figu
re 3
(
b
)
sho
w
ed
that the lo
cat
i
on of p
o
int A match
ed the
bottom very
well b
u
t B was o
b
viou
sly shifted to
north
side. Figu
re
3(c) an
d 3(d
)
sho
w
ed that
both point C and D were
locate
d nea
rby the bottom of
the curve, but
their locatio
n
s
we
re
shifted to north sid
e
as well.
This
phe
nom
enon
reve
als that if the slope of
both
side
s i
s
symmetric, th
e p
o
ints i
n
axile-line
will
matche
d wit
h
deep
est b
o
ttom nicely su
ch a
s
p
o
int A, or else it i
s
impo
ssible
to
match with each other
precisely
and the shifted di
rection
will
be f
o
rward to that side with lower
gradi
ent, su
ch as poi
nt C and D.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Applicatio
n of Rem
o
te Sensing Axi
s
Lin
e
Method in
Xiaom
iaohon
g Cre
e
k
(Yan
Gu)
3327
Although the
bias b
e
twe
e
n
axile-line an
d thalwe
g
is i
nevitable, it still has the po
tential to
figure
out th
e ch
ang
e a
n
d
evolution
of tidal cr
ee
ks be
cau
s
e
of the pri
m
a
r
y trend
can
be
expre
s
sed cl
early by axile-line meth
od.
4. Results a
nd Discu
ssi
on
4.1. The Historic Ev
olution of Xiaomiaohong Cr
eek
The axile-lin
e
s
we
re gen
erated
a
u
tomat
i
cally
o
ne
by
one
ba
sed
o
n
the
prog
re
ss
what
we have mentioned
before.
The
general trend
of Xi
aomiaohong creek
is
NWW-SEE di
rect
ion,
and mu
ch mo
re su
bo
rdin
ate bran
ch
es d
e
vel
ope
d at upper
se
ction, seem
s
claw-shape
d.
In orde
r to m
a
ke th
e evolu
t
ion of Xiaom
iaoho
ng cl
ea
rly, this cre
e
k
wa
s segme
n
ted into
3 part
s
, whi
c
h are u
ppe
r, middle an
d d
o
wn
se
ction
s
. Three ve
rtical lines
(L1,L
2
) was d
e
line
a
ted
as b
ound
ari
e
s of tho
s
e t
h
ree
pa
rts,
who
s
e lo
ngit
ude a
r
e 1
2
1
°
25
′
36
″
E a
n
d
121°
34
′
10
″
E
respe
c
tively (Figure 4).
Figure 4. Cha
nge
s of Axis Line
s in Xiao
miaoho
ng fro
m
1979 to 20
09
Upp
e
r
se
ctio
n is thou
ght
of as the
head
er of Xi
aomiao
hon
g
cre
e
k
and ti
ny bra
c
he
s
develop
ed very well in this
part. It will be helpful
to an
alyze the cha
nge of wate
r catchment fro
m
outsid
e
of
slu
i
ce
gate. Sim
ilarly, in mid
d
l
e sectio
n flo
w
in
put from
an a
d
jacent
river will
be
al
so
easily di
scussed.
4.1.1. The Ax
ile-line Chan
ge of Up
per
Section
As dete
r
mine
d above, the
uppe
r sectio
n is lo
cate
d
at we
st of vertical lin
e L1
and it
s
evolving ch
aracteri
stic fro
m
1979 to 20
03 is a
s
follo
ws:
Figure 5. The
Axile-line Ch
ange of Uppe
r Section
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3323 – 33
30
3328
(1) T
he lengt
h cha
nge of u
pper
se
ction
The va
riation
of len
g
th i
s
i
n
clin
ed to
be
elong
ate fro
m
197
9 to
20
03 ex
cept
a
backward
sho
r
tenin
g
in 1989.Th
e total increa
sed l
ength is 2
899
m from 620
2
m
of 1979 to 9101m of 20
03,
the increa
sed
length occup
i
ed 47%
percentage to the
length of 1979.
In terms of the chang
e of
the ch
annel
of uppe
r section is severe, the orig
i
n
al point is v
e
ry stabl
e an
d almo
st ha
s no
cha
nge, that
mean
s the di
stan
ce bet
we
en ori
g
inal
p
o
i
nt to its hori
z
ontal point lo
cated in
L1 h
a
s
no ch
ang
e as well, mea
n
w
hile, the len
g
th of axile-
line be
came l
onge
r, so it wa
s indi
cate
d that
the curvatu
r
e
of this se
ction
has be
en in
crea
sed.
(2) T
he swing
characte
ri
stic of uppe
r se
ction.
Acco
rdi
ng to the morp
holo
g
y of axile-line in
the uppe
r se
ction, a vertical lin
e (L
u) wh
ose
longitud
e
is 1
21°2
4
′
57
″
E was d
e
fined a
s
the bo
und
a
r
y to interp
ret
the swi
ng of
uppe
r sectio
n.
We found th
at the axile-line at the west of
Lu was inclined to north si
de a
nd the maximum
distan
ce m
o
ved wa
s a
pproxim
ate 724
m, a backward phen
ome
n
on
occu
rred
at east si
de
of Lu
,the gene
ral
moving direct
ion of axile-lin
e is fo
rward to south
side
and maximu
m distan
ce
swun
g
wa
s abo
ut 8
60m. The m
a
in characte
ri
stic of ax
ile
-li
ne of ea
ch p
e
riod i
s
a
c
co
rded
with a
b
o
ve
trend except in 1979.
(3) T
he chan
ge of sub
o
rdi
nate bra
n
che
s
of uppe
r se
ction
Subordinate
bran
ch
es
we
re
mainly
devel
ope
d at the north
side of
axile-line in t
he upp
er
se
ction of Xi
aomiao
hon
g
cre
e
k, it
s am
ount obtai
ne
d a rapid
de
velopment from 2 b
r
an
ch
es in
1979 to
4 b
r
a
n
ch
es i
n
19
8
7
, and
contin
uing in
crea
se
d to 5 b
r
an
ch
es in
198
9. After this p
e
rio
d
,
the subo
rdi
n
a
t
e bran
che
s
started to decrease, ty
pically only 1 branch left in 199
5. The length
of
axile-line of subordinate b
r
ach
e
s was al
so varie
d
sig
n
ificantly, the average
len
g
t
h of axile-line of
bran
ch
es wa
s 3
000m
in
1989,
whi
c
h
is the
pe
ak
perio
d, an
d
an o
b
viou
s d
e
crea
se
occu
rre
d
later, for exa
m
ple only 10
00m long of that in 1993.
4.1.2. The Ax
ile-line Chan
ge of Middle
Section
As determine
d above, the middle sectio
n is lo
cated b
e
twee
n vertical line L1 an
d
L2, its
evolving ch
aracteri
stic fro
m
1979 to 20
03 is a
s
follo
ws:
Figure 6. The
Axile-line Ch
ange of Middl
e Section
(1) T
he lengt
h cha
nge of
middle sectio
n
In the middle section, the
length ch
ang
ed seld
om a
nd floated 14
200m up an
d
down,
whi
c
h mea
n
s
the curvatu
r
e
of this se
ction
almost had n
o
cha
nge.
(2) T
he swing
of axile-line of middle se
ction
Acco
rdi
ng to
the mo
rp
hol
ogy of axile
-l
ine in
th
e mi
ddle
se
ction,
a verti
c
al li
ne (Lm)
who
s
e lo
ngit
ude is
121
°2
7
′
26
″
E wa
s d
e
fined a
s
the
bound
ary to
interp
ret the
swi
ng of mi
ddle
se
ction. We
found that th
e axile-lin
e a
t
the we
st of
Lm was in
cl
ined to
so
uth
side
an
d th
e
maximum di
stance
was ap
proximate
11
50m, the
mo
st active p
e
ri
o
d
o
c
curred
be
tween
19
79
a
nd
1987. A backward phe
no
menon o
c
cu
rred at ea
st si
de of Lm, the gene
ral mo
ving dire
ction
o
f
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Applicatio
n of Rem
o
te Sensing Axi
s
Lin
e
Method in
Xiaom
iaohon
g Cre
e
k
(Yan
Gu)
3329
axile-line
is f
o
rward to n
o
r
th si
de
and
maximum di
stance
swun
g
wa
s a
bout
5
00m. Th
e m
a
in
cha
r
a
c
teri
stic of axile-line of each p
e
rio
d
is acco
rde
d
with above trend except in
1979.
(3) T
he chan
ge of sub
o
rdi
nate bra
n
che
s
of uppe
r se
ction
Subordinate bran
ch
es sta
r
ted
to devel
op at the so
uth side of a
x
ile-line in th
e middle
se
ction
of
Xi
aomiao
hon
g cre
e
k. The a
v
erage
len
g
th
of axile-lin
e
of bra
n
che
s
wa
s in
a g
r
o
w
ing
perio
d bet
we
en 198
7 an
d
1993, rea
c
he
d its pea
k val
ue in 19
93 an
d be
came
sh
rinkin
g later. T
he
gene
ral movi
ng directio
n
wa
s forwa
r
d to so
uthea
st and its m
a
ximum amplitu
de re
ached
6
00m
approximatel
y.
4.1.3. The Ax
ile-line Chan
ge of Do
w
n
Section
As d
e
termin
e
d
ab
ove, the
middle
se
ctio
n is lo
cated
at ea
st si
de
of vertical lin
e L2, it
s
evolving ch
aracteri
stic fro
m
1979 to 20
03 is a
s
follo
ws:
(1) T
he lengt
h cha
nge of d
o
wn
se
ction
In
the
down
se
ction,
the
e
v
olving
cha
r
a
c
teri
stic of le
ngth of ax
ile-l
ine wa
s provided with
a sh
rin
k
ing in
former p
e
ri
o
d
and in
crea
sed late
r, the
minimum an
d maximum
wa
s in 19
93
and
2003 respe
c
tively.
(2) T
he swing
of axile-line of down
se
ction
The axile-li
n
e
of down
se
ction shifted fr
om north to
south a
nd ba
ck n
o
rth, ba
ck so
uth
again
many
times in
hi
story pe
rio
d
. Its
swi
ng
sco
pe i
s
b
e
twe
e
n
two
ho
rizo
ntal line
s
wh
ose
latitude is 32
°
9
′
15
″
N, 32
°8
′
16
″
N
respec
tively.
Figure 7. The
Axile-line Ch
ange of Down Section
From
what
we have a
n
a
lyzed a
bov
e, we
ca
n see that th
e uppe
r se
ction of
Xiaomiao
hon
g cree
k was depo
siting
and g
r
o
w
ing
at the former p
e
rio
d
, and
subo
rdi
nate
bran
ch
es
we
re g
r
ad
ually disa
ppe
are
d
. At the sa
me
time at the south
of mid
d
le sectio
n n
e
w
sub
o
rdi
nate bran
ch
es we
re develop
ed
,
whi
c
h
re
ve
aled
mu
ch
st
rong
er hyd
r
o
d
ynamic force in
this
area.
5. Conclusio
n
In orde
r to d
e
tect the
cha
nge a
nd evol
uti
on of tidal
cre
e
k
of ra
diate sand
ri
dge
s in
histori
c
p
e
rio
d
, based on
RS image
s th
is pa
per
put
forward to a
method to ex
tract axile
-lin
es of
cre
e
ks, whi
c
h
can re
presen
t its basic m
o
r
phol
ogy. The
main re
sults
were as follo
ws:
(1) Ba
se
d o
n
gen
erate
d
axile-line
s
a
n
d
mea
s
u
r
ed
bathymetry o
f
Xiaomiaoho
ng sea
area,
we fo
un
d that the g
e
neral
tren
d of
axile-
line
wa
s not m
a
tch
e
d
with th
alwe
g accu
rately
and
had bia
s
bet
wee
n
them, in terms
of the slop
e of bo
th side
wa
s n
o
t symmetri
c
. The mu
ch la
rge
r
differen
c
e
of gradi
ent of
b
o
th
si
de
s wa
s,
the
mu
ch
shift was.
Althoug
h the
bi
as
wa
s inevit
able,
after taking
consideration of the offset distanc
e, we considered that
axile-line was still an
excelle
nt sub
s
titute to thalweg a
nd be a
b
le to help u
s
carry out the evolving anal
ysis.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3323 – 33
30
3330
(2) Ba
sed
o
n
axile-li
ne
method
and
6
scene
s
of RS im
ag
es, the
axile
-line
s
o
f
Xiaomiao
hon
g cre
e
k were
extra
c
ted. In
order to di
scuss the
chan
ge of thi
s
typical
cree
k, it
wa
s
segm
ented t
o
three
pa
rts (upp
er, mid
d
le and
do
wn se
ction
s
)
a
c
cordi
ng to it
s mo
rph
o
logi
cal
cha
r
a
c
teri
stic. The
re
sult
s sh
owed
that
the
upp
er section
was d
epo
siting
and
growi
ng
all t
he
time and sub
o
rdin
ate bra
c
hes were gradually di
sap
peared, the south
side of
middle se
ction
posse
ssed re
latively stron
ger hyd
r
odyn
a
mic force a
nd much more sub
o
rdi
nat
e bran
ch
es
were
developin
g
.
B
a
se
d
o
n
the analysi
s
we
t
h
ink
this cree
k wa
s stable and had
li
mited
imp
a
ct on the
facility of Lvsi port.
(3) Mo
re tri
a
l
s
sho
u
ld b
e
continuin
g
for
revisin
g
the
shift betwe
en
axile-line
an
d
thalweg
of cree
k, pro
bably paying
more attentio
n to
the relation betwe
en slope and
shift distance wo
uld
benefit the a
d
justme
nt of extract
axile-l
ine, so that a
more
re
p
r
e
s
entative featu
r
e line
coul
d
be
gene
rated to
guarantee m
o
re a
c
curate monitori
ng re
sults.
Ackn
o
w
l
e
dg
ements
Fundin
g
for this
study
wa
s received from
Ji
ang
su Province
research and
i
n
novation
prog
ram rem
o
te sen
s
ing
identificatio
n and anal
y
s
is of the radial sa
nd ridge topog
ra
phy
(CX1
0B_39
1
Z
). Than
ks for the help fro
m
Deli Wu.
Referen
ces
[1]
Ying W
a
ng, D
a
kui Z
h
u. Rad
i
ative sa
nd
y ri
d
ge fie
l
d
o
n
co
ntine
n
tal sh
elf
of the
yel
l
o
w
sea. Bei
jin
g:
Chin
a Envir
o
n
m
ent Scienc
e Press. 2001.
[2]
Ying
W
ang, Ho
ngch
ao Xue. R
adi
ative
s
a
n
d
y
ridg
e
fie
l
d
on
c
ontin
enta
l
sh
elf
of the
yel
l
o
w
s
ea. Be
iji
ng:
Chin
a Envir
o
n
m
ent Scienc
e Press. 2001.
[3]
Yanch
un L
i
u,
Ying Z
h
ang. T
he loc
a
l
i
zatio
n
theor
y a
nd co
rrespo
ndi
ng co
rrection met
h
o
d
s of USBL.
Marine Sc
ienc
e
. 2011; 3
5
(2): 72-7
6
.
[4]
Guohu
a Yu, P
e
id
ong
Lu. A st
ud
y o
n
the
sta
b
ilit
y of s
ubmer
gibl
e tid
a
l i
n
l
e
ts.
Acta Geogra
phic
a
Sin
i
ca
.
199
6; 51(2): 12
7-13
4.
[5]
Peng H
u
, Lia
n
You. Map Alg
e
b
ra. W
uhan: W
uha
n Univ
ersit
y
Press. 20
06.
[6]
F
r
ancis Chi
n
, Jack Snoe
yi
nk
, Cao An W
ang. F
i
ndin
g
the Medi
al Axis of a Simple Po
l
y
gon i
n
Lin
e
a
r
Ti
m
e
.
Discrete
and C
o
mput
ati
ona
l Geo
m
etry
.1999; 2
1
(3): 4
05-4
20.
[7] Olson.
An
a
l
go
rithm for
ge
ner
ating
roa
d
c
ent
erlin
es fro
m
ro
ad r
i
ght-of-w
a
y
// Proceed
in
gs
of the
1
2
th
Internatio
na
l Symp
osi
u
m on
Comp
uter-Assi
sted Cartogr
ap
h
y
. C
harl
o
tte. 1995: 11-
12.
[8]
F
ederic
o T
homas. Gen
e
ratin
g
street c
enter-
l
i
nes from
i
nacc
u
rate v
e
ctor c
i
ty m
aps.
C
a
rtog
ra
ph
y a
n
d
Geogra
phic Inf
o
rmatio
n
Scien
c
e
. 1998; 2
5
(4)
:
221-23
0.
[9]
JIN Bao-s
h
i, Z
H
OU Ba
o-hu
a, HUANG
Da-p
eng. M
e
th
o
d
of
quick
l
y
e
x
tracting
w
a
ter
are
a
informati
on
i
n
remote s
ensi
n
g im
ages
b
a
se
d o
n
ArcSc
an.
Jour
nal
of
Hef
e
i
Univ
erstiy of
T
e
chn
o
lo
gy
. 2
008;
31(
11):
174
4-17
47.
Evaluation Warning : The document was created with Spire.PDF for Python.