DIP Atsgment -f

1plain bily the Seqmentation bacd on

ne eholdng Seamentation procesting incher
oixel inencty
Seting a pixd tntnslty threcbld to padktor an tmdge iat
reqionc. pxels wsth nlencitier abore the trechdd auasianed
Tid
to one reqjon while tose belos belong to ansther. Th
mathod effctive -for
(or) dictinguithing drtteert shucturer bated on thetr rlenaha
levels in the iage
Threchold baced
tihege
thage Sogmentation :
the echoique of subdiidn
*Irage Seqmedaio oic ditinct robjet.
sub- on
segi er )
image ink Conchtuent chch Subeivicion caiied out depend
The leued of detatl to
(e) Segmentatiog
on the prblem being Sclvedl;egon Se should sto
clect o) he tof 1nteret ia an
hen the
aqoication have been ditected
we Cen define the prcbem of Seamenta
flot. et Rl epesest the enire Seatial
dasde he egon
eo sach that

Rr (omeded set for i= ba--n

Q (RiuR; ) = falie
tHere QR: ) ic a logict predicte detined ovee lhe eqions in
he Set
et Rt Yepreserst he nl cet
1S one of te Seqmerstaton techues
that qenent a biniy hose piz ele have

Thus one bit to ctore pixel inkncity

imaqe hy tt into fwo
from a
witeh
xele
Rof. eqionr bared On a threchdd valie
|O00

thetoqyam of ge

Lf we have a thyechod
Cempted as

olp ha orly two cates.of prele.

..

hen the ilenstt da stibuion of dbjectr uset backqrourel

a
Suffciently distiot s it ic poscble to
|thye hdd aplac.ble ove the entreheige. thresholdT
’Select an initial estimte of
the
mge igSto fomm to
to fom to gpap Ga
 Values my ma foy
Conpite the aege olacty
valeue of ke threshold Tas:
Conpte the new
m|tm

t Repeat step a thgh t unlil the diterente n the

predefined
value of T ic Smaller then a
Subsenent
Value .
Cagmat the imaq e as qay) 1 if fay )>T

l abot Seqmentation hased on threshdd.

|Heree, this is
aDiuc briefly he eon baced Segmeatation ?
9

Regton baced Segmertaion: he

Sirmias and Connectedto the Seed
plxel cthat au Such as deffrencer to
we Use Srilacity
citu meaiwrer
evele. we
dffrent patt of tte
irage
Top don approachs predefod Seed pxel. Gthe we
we need to
defire the plxela
pixelt (or) adony choxen
Can de fne all pizele as leed
uatl all Pixelr trn the

appronch: Seleet secd only frn objectr hof

Botam -up
nerect. Criteion
 kegion mevgjing tectoiues:
he negion merT
Tegiar that Contain One tSogle
eckrinueobe ct ty to Conbine he
ard seperate it fom
the backqyaund. "Ihere ae
techniquer
such at unlordhed alyoithm, spkt ard menge alguiltn.
To diital tmege prce sting( traver daidrg as inage idto
meeningfl' egions baced Ceetain ileiay ch as pixel
Smilaut e) homagenity > Thic techniaue aime to qaup piielr
thut share Common chaactictics , making tt uceful -for taskr
lke cject ecot on Scene arlyi.
Comgn appanches
agronches a?
Rogon Gpawing
’Spit and mage
- ’ hteached Tansfomation
Mean chift Seqmeatatio n
Gpaph bared Seqeotation:
Thus t s is al about the based Seqmentation.
3. Disuuçg io detatl the igmorpholacal opeati nc:
ca) -besion (b ) Dilator
caoston & dilaion ae funamental
n digtal fage presskg mnjpdate he shape tncture
rophalical orations uled to
of chject wtthin an
Gncion:
mophagel oqpenation hat nvdler chinking
the boundacier of ohecr tn an
a element (a Small mhi
or keund) ove the imqe. At each 'oation if all he pixele
 tunder the chuctuingelement
elerment ae ty he cerder xel vemin
1; otheuwise ,tt ic Set to 0.
Smally -fie detatl fom an
imge eparating cgectr that ale close to each other,
object boundaier.
Dation:
t fEihe opporile of entorn yand it invdes epding ory
deting he bountaier of objectr in an image.
bS a to erocion, aashucturing
Sima shuctuning element is moved ore he
elementis
irge Tf at leact one pixel ukdor the thuckuing elereit
Pxe utder
,the canler plrel pixcl becones '',, sthise it emaint o.
Dilation is uce fal -for enhercinq the feahresin of
CmalU
hooges Coecting boken thuckurer and Hlng
key differerces:
Erotioo trd: to meduce the Sige of cbect: and break
Connecionc.
tends to
to inrease the Size of objed and
t Dilation
lose H them.
|plication si
ntauka ike
*Srocion s emplayed tntask bke naise reductions Sepalatirg
object , sthiening shuchures
Dlatio is used aks, bke
n teit ike fling gps jorinq
Prorrtnent!
brken
pat of obfetr ad making objetr more

Thuc; tbic is all about hansfom ?

4- Gplain atout morpholgx bit- on- nitc
The
te
hit or mics trentom
a
mohdogl
operalioo that can be uced to look for paiculay pateunt of
foreqyound and backqound preli tage Tt tr actally
Fe baie cperntion of biry moyphdogy Snee
Snce aeee almost
all the othr brany mortoageal operatorr Can be deied fem
ht.
The uctuing element ured in the ht rand mi
cutt eiteklon to the type tat has betn idhoduced for
Gih
erocioo diltion , io hat it Can Contain bitt foreqrourd ad
|backqreund xde nalhu han juct foreqrund pixeli , lebott
ones Zeos

|f1q -1: Ganle of exteded type of chuchuing element ued to ht

OY- nmics operations .

)
1

couctuirg lemedt uted for Coner -frdirg tn

in
far souting Nt
the hit-ad-mitr trancfoum,
all the Same elemet!
different amount ae
otated
of the hit- ad-mic hit-or-micr baset ight argle
Conye Coner detect on a Shrple binnuy tmage. Nole ththe
doteor ir ath Senciive
* Tlushater Some elemen that Can be used for
Vaious -featirel.
Some apicationr of the hit-ard- miss tranrfom
42s ued to locote 1solted pont n a
 uced to locate the erd pootr on a ray
N that this ottionr o far hit- tko skeletoo.
ardmicr
veained partes
ooo
3b
o 1

Ba, sb, a3c ae the keee used to locate he tiple point

a (keleton. tvee chuchuving elenent muct be n
o al enientatio So twewe hit-ad mics pases aue
neqined
tleuchnte too of thee apllcatint an image ckeleton
Wee

This
tn aailae yto (he
tile point Imige abovey
except of ocurde that a
Staut hiçe point Ed poiot of ittrent cfuctwing elemest
skeleton ued foy kit-andmise
cperaton
Thus's bit -or-misr t -d mis hransfom is use ful for dole

cing Speciic pallunr e shepe in an image that mtch the

chactig cdementt used in the process.