Pattern Recognition Full Notes Upto Mid
Pattern Recognition Full Notes Upto Mid
tnto
plaing ojecto into cossect clas
dabed on meciouSernent| featue of cofect
8 bit lmage n captured - a8 bits
Sae
rnenty
times
that uae
Pixel
OCUSs) irragfside
matlabk= histycU);
Imshouoc),
ag cf pixelo,
oange the
’ f th hitoqrarn cocOvess cll the
bett e
quality of inage io
act. ef.takina Taus
au dota
st can Ue defintd
action cuoed On caf egoy f docto,
Decision
-Cost
Roet PsoceDBlng atjostrnent of context
Cansftcot ionAostnAoetrnert of missing
etiactton
-featoreb
Feote
Segmerttotion)>baned th reshold
Sensing
Ije (Rasdata)
-’ Jomage procosing o Yequised due to
) better human pecaption
i) in undustrial aitononOub prOcUbDeb;
LosS Lossless
Redondong Yermooed)
light htah
dark Cortsast
(onder (ooercontroContxas
-esposed) exposed)
lectronic
S4nthatc imageo by conp
Bcgntees
see ctton
int (ocleoo)
corei)
Lwtne2
iq -1 Con6ider follousing
tiwo images
|25
5 16 a5
J200 Bo |BO
Ro0 60 6
0,o
256 299
50
t= f 6a a55
1 2
35
find
A+B
22
5 4 3
AxB
2
Lé 2.
LOoI
ANDoR,AND
Pefosn
AND=
6-%)
NOT
Lo oo
3x3
s 2
ay
Oo-00,
3.
.
Scale
S=3y Sy iA
S=
s' = 6Tf =
an 1 o
= l2, 4
Mio OY
Sheary o
30 o2 2 7
¿q-5-Show osthagonal
2
2 2 A.A' =I
2
2 2
2 -2 -2 2
I.2
a22
2 2I
-2+4 -
+ 4 +4 2. +2 -4
-4+ 2 t 2
2t2 -4
-24 -2 -G t 2 +2
H
Ca-6 ?- éhouo the ing
Eg-6
be econstoucod 6ing foloig
boni inageo
H ) -H= (-)
H =
Hy
-|
ty t Hi
[minos
devatlon
distostin
- Cur er e
MON
Nase Anp
=20 log,e
logo L-²
-) m
M-\ N
2 1
tB2 2 2
Sio e Ca, u)
2
i . q t 44 l41t4+|
|2.24d B
9626 dR
I22 200
Eg-2: 229 226 ’ 8 -bt
Ja.6 JA-6
34
5.4
4
O30| O34
O38 o42
7. 6-42
t o 42 O30
lo-a)
etretehind, hich is
contrat
spectol foóm df image qality oy alteing
hana
-level poeber
ong
(rom each gay
) coUtsoct a
=60 o 5
50 40
6
+ |20
48
(i20 126
|r32 l68|
S
othes Ise withot back
R, - 3, Ri -
) 22
ito olo ary mid -level
Bit plane Sicin hi h -bv
8
Pesform histogsom stoetchng n &x Leel gey
jma, he gray Qeel- dist of ohc
Umag
2 3 4 5 6
20 20
Vmin = 2
S-2
Smin :0
Smo
=&333[r-2)
| 2 34 5 6
Y=3
S=a33 =2 20 O I9
RO 20. 0
Loplacian
Ohat wll) be effect of Gaaesian filter ^
elemert
GX6
put it on crtiex ManRina
S0
ÇoY ool
bo S0 So
50
--400 4 560
TR
Cio -Sin
to cas ou
the tonsfoa
Chioe
in X & Y with factor
Sheoin9
extaten on
Pesfos bond asy
with help of
C 1
A - (A AB
one etement
doesn't
-0-o0
BCA)
.tetoggam qaigatbn - confrm
On 8x8 -’8-ouel
Pesfos n istogeiam eqzn
in table -1
2 3 4 S 6 N
2. |2 4 22
abt
Stsetching
Com x L-1
TOm ofto el TOT aL
|.46%
28 3-062 ~3
2
3 2
42
12
16 58
G8
2 64
’8-ouel
teg -2 esfourm héstosam eqn
3
3
22
2
Cum
Total
3 Q33
2
3o88
3 2
5"44
2
2
3
2
2
4x4
CUr
O4345
|"3125
2 2.
Q"18A n 2
2.
3
3
5
G-125
|4
2
2 3
specificot iÝn:
4o Histogam H S
2
Cappes value)
Eg
s5
Pesfor. histoo am epecifebin 8- louel
in. t¡lle
umag
2 3 4 5 6 4
16 4 2
8 2 12
table-z
3
20 20 6
Comm
Roun d
Total
2
20
20 4
20 0 375
G"125
ttncl
S Cmap Ping
HCrnap OF tasget)
Gay level of ea) L
2
4
2
2
5
6
Appy histogam peciflaton to gwen image
Assurme tosget histogramn & sh
hese
histo gr ene
gesute
3 4
2. 3 45
3 4
22 2 3 6
3 2
2
Value
3
2
O434
3 2
3 I13125
&.
4
12,
4
=(o,
2) (o)(o,1)+
A5 )+(0,
)-0 (,6) (o,
(Ocation
c) , = o)(, +) =
=o,) ))(o, (o,o)+
10coti
cm
1's check
2
2
B
AD
sUctus
ing with
elarmert
proces9anaustie
dddition i)9ector
Sobstuttion )
=O match No
3 Dilation -18- &g
match Ofoll
Struusing
darme
CAs) (osung 4)
(AB) Opening 3)
object of
Stsuctuse Se
¬ 3hape
(A
8)DUution )
) AG
match NO
match Sore
-2 match Foll O
b
ignthess 9isen
pKlgboundosy emoveo
osioni
Openngt
Kernoveo
boordcsy of cojets
followed by eso(n
AB diution
E(D(A))
=AOB =
D(A) A
’ Bit 2 3 sty
plane
Slicing
LSB
middlo
MSB