F (u )
f ( x)e j 2 ux dx
Given Fourier transform of a function F(u), the inverse Fourier transform can be used to
obtain, f(x), by:
f ( x)
F (u )e j 2 ux du
Note: F(u), is the frequency spectrum, where, u represents the frequency, and f(x) is the
signal where x represents time/space.
j 1
Prepared By: Dr. Hasan Demirel, PhD
Time/Space domain
Sine wave
-1
-2
0.2
0.4
0.6
0.8
1.2
1.4
1.6
1.8
300
250
Delta function
200
Frequency domain
150
100
50
0
20
40
60
80
100
120
0.5
0.4
Time/Space domain
0.3
Gaussian
0.2
0.1
0
10
15
20
25
30
35
40
45
50
6
5
Frequency domain
Gaussian
3
2
1
0
50
100
150
200
250
1.5
Time/Space domain
Sinc function
0.5
-0.5
-1
-0.8
-0.6
-0.4
-0.2
0.2
0.4
0.6
0.8
6
5
Frequency domain
Square wave
3
2
1
0
-100
-50
50
100
F (u )
f ( x )e
M
x 0
The inverse Discrete Fourier Transform can be used to calculate f(x), by:
M 1
f ( x) F (u )e j 2 ux / M
u 0
Note: F(u), which is the Fourier transform of f(x) contains discrete complex quantities and
it has the same number of components as f(x).
e j cos j sin
Prepared By: Dr. Hasan Demirel, PhD
1
F (u )
M
M 1
f ( x)[cos 2 ux / M j sin 2 ux / M ]
x 0
The Fourier Transform generates complex quantities. The magnitude or the spectrum of
the Fourier transform is given by:
1/ 2
R(u) is the Real Part and
2
2
F (u ) R (u ) I (u )
The Phase Spectrum of the transform refers to the angles between the real and imaginary
components and it is denoted by:
I (u )
(u ) tan
R
(
u
)
P(u) F (u ) R 2 (u ) I 2 (u )
2
The power spectrum can be used, for example to separate a portion of a specified
frequency (i.e. low frequency) power from the power spectrum and monitor the effect.
Typically used to define the cut off frequencies used in lowpass and highpass filtering.
1
F (u, v)
MN
M 1 N 1
j 2 ( ux / M vy / N )
f
(
x
,
y
)
e
x 0 y 0
u=0, 1, 2, , M-1, and v=0, 1, 2, , N-1 and the inverse 2D Discrete Fourier Transform
can be calculated by:
M 1 N 1
f ( x, y )
u 0 v 0
1/ 2
Magnitude/Fourier
Spectrum
I (u, v)
(u, v) tan
R(u, v)
1
Phase Spectrum
Power Spectrum
Centered F(u,v)
Centered + Log
Transformed F(u,v)
Origin = 0,0
f(x,y)
F(u,v)
f(x,y)
F(u,v)
f(x,y)
F(u,v)
One of the properties of the 2D
FT is that if you rotate the
image, the spectrum will rotate
in the same direction
f(x,y)
F(u,v)
The sum of 2 sine functions, each
in opposite (vertical and
horizontal) direction.
1
F (0,0)
MN
F(u,v)
f(x,y)
f ( x, y)
x 0 y 0
f(x,y)
F(u,v)
f(x,y)
(Average image)
F(u,v)
(DC Component)
Prepared By: Dr. Hasan Demirel, PhD
F(u,v)
The lines in an image often
generate perpendicular lines
in the spectrum
f(x,y)
F(u,v)
The sloped lines in the
spectrum are due to the sharp
transition from the sky to the
mountain
Given the filter H(u,v) (filter transfer function) in the frequency domain, the Fourier
transform of the output image (filtered image) is given by:
Step (3)
The filtered image g(x,y) is simply the inverse Fourier transform of G(u,v).
g ( x, y) 1G(u, v)
Step (4)
0
H (u, v)
1
if (u, v) ( M / 2, N / 2)
otherwise
This filter will set F(0,0) to zero and leave all the other frequency components. Such a
filter is called the notch filter, since it is constant function with a hole (notch) at the origin.
1
H (u, v)
0
if D(u, v) D0
if D(u, v) D0
1
H (u, v)
1 [ D(u, v) / D0 ]2 n
H (u, v) e
D 2 ( u ,v ) / 2 D02
Original image
Ideal LP Filtered
Butterworth LP Filtered
Gaussian LP Filtered
Smoothing/blurring can be achieved by attenuating a specified range of highfrequency components in the frequency domain.
Given the Fourier transformed image F(u) , the filtered image G(u) can be
obtained by:
1
H (u, v)
0
if D(u, v) D0
if D(u, v) D0
Cutoff frequency
PT P(u, v)
u 0, 1, ..., M 1 , v 0, 1, ..., N 1
u 0 v 0
a 100 P(u, v) / PT
u v
u radius ( Do )
, v radius ( Do )
Original Image
Fourier Spectrum
15 pixels , a = 92%
80 pixels , a = 98%
230 pixels ,
a = 99.5%
Prepared By: Dr. Hasan Demirel, PhD
cutoff=15,
r.d. power=5.4%
cutoff=80,
r.d. power=2%
The corresponding process in the spatial domain with regard to the convolution
theorem is given by:
g ( x, y) h( x, y) f ( x, y)
h(x,y) in the spatial domain, is the inverse Fourier transform of the filter
transfer function H(u,v).
The spatial filter h(x,y) has two major characteristics:
dominant component at the origin
Concentric circular components about center component.
Prepared By: Dr. Hasan Demirel, PhD
inverse DFT
=
circular components: causes ringing effect
1
H (u, v)
2n
1 [ D(u, v) / D0 ]
D0 =5, n=2
D0 =5, n=3
D0 =5, n=4
cutoff=15,
order=2
cutoff=80,
order=2
H (u, v) e
D 2 ( u ,v ) / 2 D02
cutoff=15
cutoff=230
cutoff=80
H hp (u, v) 1 H lp (u, v)
Where Hlp(u,v), is the transfer function of the corresponding lowpass filter.
0
H (u, v)
1
Perspective plot
of IHPF
if D(u, v) D0
if D(u, v) D0
Image representation
of IHPF
Cross section of
IHPF
Prepared By: Dr. Hasan Demirel, PhD
D0 =15
D0 =30
D0 =80
1
H (u, v)
2n
1 [ D0 / D(u, v)]
Perspective plot
of BHPF
Image representation
of BHPF
Cross section of
BHPF
Prepared By: Dr. Hasan Demirel, PhD
D0 =15, n=2
D0 =30, n=2
D0 =80, n=2
H (u, v) 1 e
Perspective plot
of BHPF
D 2 ( u ,v ) / 2 D02
Image representation
of BHPF
Cross section of
BHPF
Prepared By: Dr. Hasan Demirel, PhD
D0 =15
D0 =30
D0 =80