This preview shows pages 1–11. Sign up to view the full content.
1
IEG 4190 Lecture Notes 5;
Lecturer: Jianzhuang Liu
Outline:
1.
Grayscale Modification
2.
Highpass Filtering
3.
Homomorphic Processing
4.
Adaptive Processing
5.
Lowpass Filtering
6.
Median Filtering
7.
Edge Detection
8.
Motion Estimation
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document 2
Grayscale Modification
Grayscale modification:
•
Increasing contrast of original degraded images to make them look
sharper.
•
Recording medium may not have enough dynamic range.
•
Contrast reduction due to low light, mist, or cloud cover.
Depends on applications:
•
Gamma correction.
•
Manual gray scale modification.
•
Histogram modification.
3
Grayscale Modification, cont’d
Gamma correction: The electrooptical transfer characteristic of the
CRT is inherently nonlinear. It is necessary to compensate for the
CRT nonlinearity by introducing a compensating nonlinearity, usually
referred to as gamma correction.
ANSI standard:
CRT:
γ
⎥
⎦
⎤
⎢
⎣
⎡
+
=
099
.
1
)
099
.
0
(
r
T
V
L
for
1
0812
.
0
≤
≤
r
V
5
.
4
r
T
V
L
=
f
o
r
0812
.
0
0
≤
≤
r
V
r
V
is the video signal level;
T
L
is the light output
Camera
099
.
0
099
.
1
)
/
1
(
−
×
=
C
C
L
V
for
1
0812
.
0
≤
≤
C
L
C
C
L
V
×
=
500
.
4
f
o
r
0812
.
0
0
≤
≤
C
L
C
V
is the video signal output;
C
L
is the light input,
2
.
2
=
.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document 4
Grayscale Modification, cont’d
5
Grayscale Modification, cont’d
Manual grayscale modification
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document 6
Grayscale Modification, cont’d
Histogram: number of pixels as a function of image intensity.
7
Grayscale Modification, cont’d
Examples of images and their histograms
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document 8
Grayscale Modification, cont’d
Histogram modification:
Input histogram:
)
(
f
p
Desired output histogram:
)
(
g
p
d
, typically smooth and spread out.
()
gT
f
=
such that
d
p
gp
g
→
;
T
: monotonically nondecreasing.
Compute the cumulative histogram:
∑
=
+
−
=
=
f
k
f
p
f
P
k
p
f
P
0
)
(
)
1
(
)
(
)
(
∑
=
+
−
=
=
g
k
d
d
d
d
g
p
g
P
k
p
g
P
0
)
(
)
1
(
)
(
)
(
Choose the transformation so that for each
f
, choose
g
so that
)
(
g
P
d
is as close to
)
(
f
P
as possible.
9
Grayscale Modification, cont’d
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document 10
Grayscale Modification, cont’d
()
pf
d
pg
?
T
=
M
M
Pf
d
Pg
'
T
'
TT
⇐
Finding
is not easy.
Convert the problem to an easier one where finding
' is easy.
This is the end of the preview. Sign up
to
access the rest of the document.
This note was uploaded on 10/13/2009 for the course CS 5150 taught by Professor Xulei during the Spring '09 term at University of Central Arkansas.
 Spring '09
 xulei

Click to edit the document details