L-11 (Digital ImageProcessing-Image Enhancement in Frequency Domain)

COL LEGE OF E N GI N E ER I N G ROOR K EE
Established in 1998
Digital Image Processing (TCS-071)
Session: 2020-21
Digital Image Processing

COURSE OBJECTIVE
The aim of this course is,

 Describe and explain basic principles of digital image
processing.
 Design and implement algorithms that perform basic image
processing (e.g. noise removal and image enhancement).
 Design and implement algorithms for advanced image
analysis (e.g. image compression, image segmentation).
 Assess the performance of image processing algorithms and
systems.

Unit 2
Image Enhancement in Frequency Domain
Fourier Transform and the Frequency Domain, Basis of

Filtering in Frequency Domain, Filters – Low-pass, High-
pass; Correspondence Between Filtering in Spatial and
Frequency Domain; Smoothing Frequency Domain
Filters – Gaussian Lowpass Filters; Sharpening Frequency
Domain Filters – Gaussian Highpass Filters;
Homomorphic Filtering.

Image Enhancement in the
Frequency Domain

Objective
• Basic understanding of the frequency

domain, Fourier series, and Fourier
transform.
• Image enhancement in the frequency
domain.

Fourier Series
Any function that periodically
repeats itself can be expressed
as the sum of sines and/or
cosines of different
frequencies, each multiplied
by a different coefficients.
This sum is called a Fourier
series.

Fourier Series

g (t )  a0   an cos nw0t  bn sin nw0t
n 1
t 0 T0
1
a0 
T0  g (t )dt
t0
t 0 T0
2
an 
T0  g (t ) cos nw t dt
t0
0
t 0 T0
2
bn 
T0  g (t ) sin nw t dt
t0
0

Fourier Transform
A function that is not periodic but the area under

its curve is finite can be expressed as the integral
of sines and/or cosines multiplied by a weighing
function. The formulation in this case is Fourier
transform.

Continuous One-Dimensional Fourier Transform
and Its Inverse

 f ( x )e
 j 2ux
F (u )  dx

 Where j  1
 F (u )e
j 2ux
f ( x)  du

• (u) is the frequency variable.
• F(u) is composed of an infinite sum of sine and
cosine terms and…
• Each value of u determines the frequency of its
corresponding sine-cosine pair.
and Its Inverse
Example
Find the Fourier transform of a gate function (t)
defined by
 1
1 x   u 
 ( x)   2 F (u )   sinc  
0
1
x   2 
 2

x u
- 0.5 0.5
2 4
Digital Image Processing  
and Its Inverse

Discrete One-Dimensional Fourier Transform and
Its Inverse
• A continuous function f(x) is discretized into a sequence:
{ f ( x0 ), f ( x0  x), f ( x0  2x),..., f ( x0  [ N  1]x)}
by taking N or M samples x units apart.

Its Inverse
• Where x assumes the discrete values (0,1,2,3,…,M-1)

then
f ( x)  f ( x0  xx)
• The sequence {f(0),f(1),f(2),…f(M-1)} denotes any

M uniformly spaced samples from a corresponding
continuous function.

Its Inverse
M 1 x
1  j 2u
F (u ) 
M
 f ( x )e M u =[0,1,2, …, M-1]
x 0
M 1
1  x x
F (u ) 
M
 f ( x) cos 2u
 M
 j sin 2u 
M
x 0
M 1 u
j 2 x
f ( x)   F (u )e M
u 0
x =[0,1,2, …, M-1]
Its Inverse
• The values u = 0, 1, 2, …, M-1 correspond to samples of

the continuous transform at values 0, u, 2u, …, (M-
1)u.
i.e. F(u) represents F(uu), where:
1
u 
Mx
• Each term of the FT (F(u) for every u) is composed of the
sum of all values of f(x)

Its Inverse
• The Fourier transform of a real function is generally

complex and we use polar coordinates:
F (u )  R (u )  jI (u )
j ( u )
F (u )  F (u ) e
F (u )  [ R 2 (u )  I 2 (u )]1/ 2
 I (u ) 
 (u )  tan  1

 R (u ) 
Its Inverse
|F(u)| (magnitude function) is the Fourier spectrum of

f(x) and (u) its phase angle.
• The square of the spectrum

2
P (u )  F (u )  R 2 (u )  I 2 (u )
is referred to as the Power Spectrum of f(x)

(spectral density).

Discrete 2-Dimensional Fourier Transform
• Fourier spectrum: 
F (u , v)  R (u, v)  I (u , v)
2 2

1/ 2
 I (u , v) 
• Phase:  (u, v)  tan  1

 R (u , v ) 
2
• Power spectrum: P (u , v)  F (u , v)  R 2 (u , v)  I 2 (u , v)

Its Inverse

Time and Frequency Resolution and Sampling
Fmax = 100 Hz
What is the sampling rate (Nyquist)?
What is the time resolution?
What is the frequency resolution?
What if the sampling rate is higher
than the Nyquist sampling rate?
What if we take samples for two
seconds with the Nyquist sampling
rate?

1 second
Discrete Two-Dimensional Fourier Transform and
Its Inverse
M 1N 1 x y
1  j 2 ( u v )
F (u , v) 
MN
  f ( x, y )e M N
x 0 y 0
M 1N 1 u v
j 2 ( x  y )
f ( x, y )   F (u , v)e M N
u 0 v 0
 2
F (u , v)  R (u , v)  I (u , v) 2

1/ 2
Fourier Spectrum
Discrete Two-Dimensional Fourier Transform and Its
Inverse
M 1N 1
1
F (0,0) 
MN
  f ( x, y )
x 0 y 0
F(0,0) is the average intensity of an image

Discrete Two-Dimensional Fourier Transform and Its
Inverse
Use Matlab to generate the above figures

Frequency Shifting Property of the Fourier Transform
If
g (t )  G ( )
then
j 0 t
g (t )e  G (  0 )
x y
j 2 ( u 0  v0 )
f ( x, y ) e M N
 F (u  u0 , v  v0 )
Frequency Shifting Property of the Fourier Transform

Basic Filtering in the Frequency Domain using Matlab
function Normalized_DFT = Img_DFT(img)
img=double(img); % So mathematical operations can be conducted on
% the image pixels.
[R,C]=size(img);
for r = 1:R % To phase shift the image so the DFT will be
for c=1:C % centered on the display monitor
phased_img(r,c)=(img(r,c))*(-1)^((r-1)+(c-1));
end
end
fourier_img = fft2(phased_img); %Discrete Fourier Transform
mag_fourier_img = abs(fourier_img ); % Magnitude of DFT
Log_mag_fourier_img = log10(mag_fourier_img +1);
Max = max(max(Log_mag_fourier_img ));
Normalized_DFT = (Log_ mag_fourier_img )*(255/Max);
imshow(uint8(Normalized_DFT))
Basic Filtering in the Frequency Domain
1. Multiply the input image by (-1)x+y to center the transform

2. Compute F(u,v), the DFT of the image from (1)
3. Multiply F(u,v) by a filter function H(u,v)
4. Compute the inverse DFT of the result in (3)
5. Obtain the real part of the result in (4)
6. Multiply the result in (5) by (-1)x+y
Filtering out the DC Frequency Component

Filtering out the DC Frequency Component
Notch Filter
0 if (u , v)  ( M / 2, N / 2)
H (u , v)  
1 otherwise
Low-pass and High-pass Filters
Low Pass Filter attenuate
high frequencies while
“passing” low frequencies.
High Pass Filter

attenuate low frequencies
while “passing” high
frequencies.




Smoothing Frequency Domain, Ideal Low-pass Filters
1 if D(u , v)  D0
H (u , v)  
0 if D(u , v)  D0

D(u, v)  (u  M / 2)  (v  N / 2)
2

2 1/ 2

M 1N 1
  F (u, v)
2
Total Power PT 
u 0 v 0
The remained percentage  

power after filtration   100   F (u, v) / PT 
Digital Image Processing u v 
fc =5
 = 92%
fc =15 fc =30
 = 94.6%  = 96.4%
fc =80 fc =230
 = 98%  = 99.5%
Cause of Ringing

Project #5, Ideal Low-pass Filter
Implement in Matlab the Ideal low-pass filter in the
following equation.
1 if D(u , v)  D0
H (u , v)  
0 if D(u , v)  D0

D (u , v)  (u  M / 2)  (v  N / 2)
2

2 1/ 2
1. You must give the user the ability to specify the cutoff
frequency D0
2. Calculate the The remained percentage power after
filtration.
3. Display the image after filtration
4. Use Elaine image to test your program
Smoothing Frequency Domain, Butterworth Low-pass
Filters
1
H (u , v) 
1  D (u , v) / D0 
2n

Filters
Butterworth Low-pass Radii= 5
Filter: n=2
Radii= 15
Radii= 30
Radii= 80 Radii= 230

Filters

Smoothing Frequency Domain, Gaussian Low-pass
Filters
 D 2 ( u ,v ) / 2 D02
H (u , v)  e
Filters
Gaussian Low-pass Radii= 5
Radii= 15
Radii= 30
Radii= 80 Radii= 230

Filters

Filters

Filters

Sharpening Frequency Domain Filters
Hhp(u,v) = 1 - Hlp(u,v)

Sharpening Frequency Domain Filters

Sharpening Frequency Domain, Ideal High-pass Filters
0 if D(u , v)  D0
H (u , v)  
1 if D(u , v)  D0
Sharpening Frequency Domain, Butterworth High-
pass Filters
1
H (u , v) 
1  D0 / D (u , v)
2n

Sharpening Frequency Domain, Gaussian High-pass
Filters
 D 2 ( u ,v ) / 2 D02
H (u, v)  1  e
Homomorphic Filtering

Homomorphic Filtering

Convolution
g(t) = k(t)*h(t)
k(t) h(t) g(t)
G(f) = K(f)H(f)
* is a convolution operator and not multiplication

M 1
1
g ( t )  k ( t ) * h( t ) 
M
 k (t  m)h(m)
m 0
k(t)
h(t)
3
2
1
1
t t
-1 Digital Image Processing 1 2 3
Convolution
h(m)
M 1
1
1
g (0) 
M
 k (0  m)h(m)
m 0
m
g(t)
-1
k(-m)
m
Convolution
h(m)
1
g(t)
m
-1
k(1-m)
m
Convolution
h(m)
1
g(t)
m
-1
k(2-m)
m
Convolution
h(m)
1
g(t)
m
-1
k(3-m)
m
Convolution
h(m)
1
g(t)
m
-1
k(4-m)
m
Convolution
h(m)
1
g(t)
m
-1
k(5-m)
m
Convolution
h(m)
1
g(t)
m
-1
k(6-m)
m
2-Dimensions Convolution
M 1 N 1
1
f ( x , y ) h( x , y ) 
MN
  f (m, n)h( x  m, y  n)
m 0 n  0

Correlation
h(t) g(t) = k(t)h(t)

k(t) g(t) G(f) = K(f)* . H(f)
M 1
1
g ( t )  k ( t )  h( t ) 
M
 k (t  m)h(m)
m 0
k(t) g(t)
h(t)
2 2
1 1
t t
-1
Correlation

Convolution

Convolution

Convolution

Convolution

THANK YOU!

L-11 (Digital ImageProcessing-Image Enhancement in Frequency Domain)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

L-11 (Digital ImageProcessing-Image Enhancement in Frequency Domain)

Uploaded by

Copyright:

Available Formats

COL LEGE OF E N GI N E ER I N G ROOR K EE

Digital Image Processing (TCS-071)

Digital Image Processing

The aim of this course is,

Digital Image Processing

Fourier Transform and the Frequency Domain, Basis of

Digital Image Processing

Digital Image Processing

• Basic understanding of the frequency

Digital Image Processing

Digital Image Processing

Digital Image Processing

A function that is not periodic but the area under

Digital Image Processing

Digital Image Processing

• A continuous function f(x) is discretized into a sequence:

{ f ( x0 ), f ( x0  x), f ( x0  2x),..., f ( x0  [ N  1]x)}

by taking N or M samples x units apart.

Digital Image Processing

• Where x assumes the discrete values (0,1,2,3,…,M-1)

• The sequence {f(0),f(1),f(2),…f(M-1)} denotes any

Digital Image Processing

• The values u = 0, 1, 2, …, M-1 correspond to samples of

• The Fourier transform of a real function is generally

|F(u)| (magnitude function) is the Fourier spectrum of

• The square of the spectrum

is referred to as the Power Spectrum of f(x)

Digital Image Processing

Digital Image Processing

Digital Image Processing

Digital Image Processing

F(0,0) is the average intensity of an image

Digital Image Processing

Use Matlab to generate the above figures

Digital Image Processing

Digital Image Processing

1. Multiply the input image by (-1)x+y to center the transform

Digital Image Processing

High Pass Filter

Digital Image Processing

Digital Image Processing

Digital Image Processing

Digital Image Processing

Digital Image Processing

The remained percentage  

Digital Image Processing

Digital Image Processing

Radii= 80 Radii= 230

Digital Image Processing

Digital Image Processing

Radii= 80 Radii= 230

Digital Image Processing

Digital Image Processing

Digital Image Processing

Digital Image Processing

Digital Image Processing

Digital Image Processing

Digital Image Processing

Digital Image Processing

Digital Image Processing

* is a convolution operator and not multiplication

Digital Image Processing

h(t) g(t) = k(t)h(t)

Digital Image Processing