Lecture 05
Image Enhancement
in Frequency Domain
Contents
In this lecture we will look at image enhancement in
frequency domain:
1-D and 2-D Fourier transforms
1-D and 2-D discrete Fourier transform
Lowpass filters: ideal, Butterworth, Gaussian
Highpass filters: ideal, Butterworth, Gaussian
Important properties of Fourier transform
Image Enhancement in Frequency Domain 2
Background
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
coefficient (Fourier series).
Even functions that are not periodic (but whose area
under the curve is finite) can be expressed as the
integral of sines and/or cosines multiplied by a
weighting function (Fourier transform).
The purpose of the Fourier transform is to represent a
signal as a linear combination of sinusoidal signals of
various frequencies.
Image Enhancement in Frequency Domain 3
Background
Image Enhancement in Frequency Domain 4
Fourier transform
The one-dimensional Fourier transform and its
inverse
• Fourier transform (continuous case)
F (u ) f ( x)e j 2ux dx where j 1 e j cos j sin
• Inverse Fourier transform: f ( x) F (u )e j 2ux du
The two-dimensional Fourier transform and its inverse
• Fourier transform (continuous case)
F (u , v) f ( x, y )e j 2 ( ux vy ) dxdy
• Inverse Fourier transform:
f ( x, y ) F (u , v)e j 2 ( ux vy ) dudv
Image Enhancement in Frequency Domain 5
Discrete Fourier transform
The one-dimensional discrete Fourier transform and its
inverse
• Discrete Fourier transform (DFT)
M 1
1 j 2ux / M
F (u )
M
f ( x )e
x 0
for u 0,1,2,..., M 1
• Inverse discrete Fourier transform (IDFT):
M 1
f ( x) F (u )e j 2ux / M for x 0,1,2,..., M 1
u 0
Image Enhancement in Frequency Domain 6
Fourier transform and frequency domain
F(u) can be expressed in polar coordinates:
F (u ) F (u ) e j (u )
1
2
where F (u ) R (u ) I (u ) 2
2 (magnitude or spectrum)
I (u )
(u ) tan 1 (phase angle or phase spectrum)
R (u )
• R(u): the real part of F(u)
• I(u): the imaginary part of F(u)
Power spectrum: 2
P(u ) F (u ) R 2 (u ) I 2 (u )
Image Enhancement in Frequency Domain 7
1D- Fourier transform-Example
Image Enhancement in Frequency Domain 8
1D Fourier transform-Examples
The transform of a constant function is a DC value only.
The transform of a delta function is a constant.
Image Enhancement in Frequency Domain 9
1D Fourier transform-Examples
The transform of an infinite train of delta functions spaced
by T is an infinite train of delta functions spaced by 1/T.
The transform of a cosine function is a positive delta at the
appropriate positive and negative frequency.
Image Enhancement in Frequency Domain 10
1D Fourier transform-Examples
The transform of a sin function is a negative complex delta
function at the appropriate positive frequency and a
negative complex delta at the appropriate negative
frequency.
The transform of a square pulse is a sinc function.
Image Enhancement in Frequency Domain 11
2D discrete Fourier transform
Discrete Fourier transform (2D-DFT)
M 1 N 1
1
F (u , v) f ( x, y )e j 2 (ux / M vy / N )
MN x 0 y 0
for u 0,1,2,..., M 1, v 0,1,2,..., N 1
Inverse discrete Fourier transform (2D-IDFT)
M 1 N 1
f ( x, y ) F (u , v)e j 2 ( ux / M vy / N )
u 0 v 0
for x 0,1,2,..., M 1, y 0,1,2,..., N 1
- u, v : the transform or frequency variables
- x, y : the spatial or image variables
Image Enhancement in Frequency Domain 12
Fourier transform-frequency domain
We define the Fourier spectrum, phase angle, and
power spectrum as follows:
1
2 2
F (u, v) R (u , v) I (u , v)
2 ( spectrum)
I (u, v)
1
(u , v) tan (phase angle)
R (u, v)
2
P (u,v) F (u , v) R 2 (u, v) I 2 (u , v) (power spectrum)
• R(u,v): the real part of F(u,v)
• I(u,v): the imaginary part of F(u,v)
Image Enhancement in Frequency Domain 13
Amplitude spectrum
Image Enhancement in Frequency Domain 14
2D DFT with different functions
Sine wave Its DFT
Rectangle
Its DFT
Image Enhancement in Frequency Domain 15
2D DFT with Different Functions
2D Gaussian Its DFT
function
Impulses
Its DFT
Image Enhancement in Frequency Domain 16
Properties of Fourier transform
Some properties of Fourier transform:
M N
f ( x, y )(1) x y
F (u , v ) (shift)
2 2
M 1 N 1
1
F (0,0)
MN
f ( x, y)
x 0 y 0
(average)
F (u, v) F * (u ,v) (conujgate symmetric)
F (u , v) F (u,v) (symmetric)
Image Enhancement in Frequency Domain 17
Properties of Fourier transform: Shift
shift
Image Enhancement in Frequency Domain 18
Properties of Fourier transform: Separable
The 2D DFT F(u,v) can be obtained by
1. taking the 1D DFT of every row of image f(x,y), F(u,y),
2. taking the 1D DFT of every column of F(u,y)
(a) f(x,y) (b) F(u,y) (c) F(u,v)
Image Enhancement in Frequency Domain 19
Properties of Fourier transform: Rotation
DFT
DFT
Image Enhancement in Frequency Domain 20
Properties of Fourier transform:
Linear combination
A
DFT
B
DFT
0.25 * A
+ 0.75 * B
DFT
Image Enhancement in Frequency Domain 21
Properties of Fourier transform: Expansion
A
DFT
B DFT
Expanding the original image by a factor of n (n=2), filling
the empty new values with zeros, results in the same DFT.
Image Enhancement in Frequency Domain 22
Filtering in frequency domain
Image Enhancement in Frequency Domain 23
Some Basic Filters
Multiply all values of F(u,v) by the filter function (notch filter):
0 if (u , v ) ( M / 2, N / 2)
H (u , v)
1 otherwise.
• All this filter would do is set F(0,0) to zero (force the average
value of an image to zero) and leave all frequency components of
the Fourier transform untouched.
Image Enhancement in Frequency Domain 24
Some Basic Filters
Lowpass filter
Highpass filter
Image Enhancement in Frequency Domain 25
Filtering in spatial and frequency domain
The discrete convolution of two functions f(x,y) and
h(x,y) of size MXN is defined as
M 1 N 1
1
f ( x, y ) h ( x, y )
MN
f (m, n)h( x m, y n)
m 0 n 0
Let F(u,v) and H(u,v) denote the Fourier
transforms of f(x,y) and h(x,y), then
f ( x, y ) h( x, y ) F (u , v) H (u , v)
f ( x, y )h( x, y ) F (u , v) H (u , v)
Image Enhancement in Frequency Domain 26
Gaussian filter
Let H(u) denote a frequency domain, Gaussian filter
function given the equation
u 2 / 2 2
H (u ) Ae
where : the standard deviation of the Gaussian curve.
The corresponding filter in the spatial domain is
2 2 2 x 2
h( x) 2 Ae
Note: Both the forward and inverse Fourier transforms of
a Gaussian function are real Gaussian functions.
Image Enhancement in Frequency Domain 27
Gaussian filter
Image Enhancement in Frequency Domain 28
Gaussian filter
One very useful property of the Gaussian function is that both it and
its Fourier transform are real valued; there are no complex values
associated with them.
In addition, the values are always positive. So, if we convolve an
image with a Gaussian function, there will never be any negative
output values to deal with.
There is also an important relationship between the widths of a
Gaussian function and its Fourier transform. If we make the width
of the function smaller, the width of the Fourier transform gets larger.
This is controlled by the variance parameter 2 in the equations.
These properties make the Gaussian filter very useful for lowpass
filtering an image. The amount of blur is controlled by 2. It can be
implemented in either the spatial or frequency domain.
Other filters besides lowpass can also be implemented by using two
different sized Gaussian functions.
Image Enhancement in Frequency Domain 29
Smoothing Frequency-Domain Filters
The basic model for filtering in the frequency domain
G (u , v ) H (u , v) F (u , v)
where F(u,v): the Fourier transform of the image to be
smoothed
H(u,v): a filter transfer function
Smoothing is fundamentally a lowpass operation in the
frequency domain.
There are several standard forms of lowpass filters (LPF).
• Ideal lowpass filter
• Butterworth lowpass filter
• Gaussian lowpass filter
Image Enhancement in Frequency Domain 30
Ideal lowpass filters (ILPFs)
The simplest lowpass filter is a filter that “cuts off” all
high-frequency components of the Fourier transform
that are at a distance greater than a specified
distance D0 from the origin of the transform.
The transfer function of an ideal lowpass filter
1 if D (u , v) D0
H (u , v)
0 if D(u , v) D0
where D(u,v) : the distance from point (u,v) to the
center of their frequency rectangle
1
D(u , v) (u M / 2) 2 (v N / 2)
2 2
Image Enhancement in Frequency Domain 31
Ideal lowpass filters (ILPFs)
Image Enhancement in Frequency Domain 32
Ideal lowpass filters (ILPFs)
Image Enhancement in Frequency Domain 33
Ideal Lowpass Filters
• As the filter radius
increases, less and less
power is removed, resulting
in less severe blurring.
• Ring is evident even the
image in which only 2% of
the total power was
removed.
Image Enhancement in Frequency Domain 34
Butterworth lowpass filters (BLPFs) with
order n
1
H (u , v)
1 D(u , v) / D0
2n
• The filter’s smooth transition between low and high frequencies
• No ringing visible in any of the image processed with BLPFs
Image Enhancement in Frequency Domain 35
Butterworth lowpass
filters
Butterworth lowpass
filters (BLPFs)
n=2
D0=5,15,30,80,and 230
• No ringing visible in any
of the image processed
with BLPFs
Image Enhancement in Frequency Domain 36
Gaussian lowpass filters (GLPFs)
D 2 ( u , v ) / 2 D02
H (u , v) e
• The GLPFs did not achieve as much smoothing as the BLPFs.
• No ringing for the GLPFs.
Image Enhancement in Frequency Domain 37
Gaussian lowpass
filters (FLPFs)
Gaussian Lowpass
Filters (FLPFs)
D0=5,15,30,80, and 230
• No ringing for the
GLPFs.
Image Enhancement in Frequency Domain 38
Additional examples of lowpass filtering
Image Enhancement in Frequency Domain 39
Sharpening frequency domain filter
H hp (u , v) 1 H lp (u , v)
Ideal highpass filter
0 if D (u , v ) D0
H (u , v)
1 if D (u , v) D0
Butterworth highpass filter
1
H (u , v )
1 D0 / D (u , v)
2n
Gaussian highpass filter
D 2 ( u ,v ) / 2 D02
H (u , v ) 1 e
Image Enhancement in Frequency Domain 40
Highpass Filters Spatial Representations
Image Enhancement in Frequency Domain 41
Ideal highpass filters
0 if D (u , v) D0
H (u , v)
1 if D (u , v) D0
• Ringing problems are evident in (a) and (b)
Image Enhancement in Frequency Domain 42
Butterworth highpass filters
1
H (u , v)
1 D0 / D(u , v)
2n
Image Enhancement in Frequency Domain 43
Gaussian highpass filters
D 2 ( u ,v ) / 2 D02
H (u , v) 1 e
Image Enhancement in Frequency Domain 44
Laplacian in Frequency Domain
The Laplacian filter
H (u , v) (u 2 v 2 )
Shift the center: Frequency
domain
M N
H (u , v ) (u ) 2 (v ) 2
2 2
Spatial domain
Image Enhancement in Frequency Domain 45
g ( x, y ) f ( x, y ) 2 f ( x, y )
where
2 f ( x, y ) : the Laplacian - filtered
image in the spatial domain
For display
purposes only
Image Enhancement in Frequency Domain 46
Additional properties of 2D Fourier
transform
Periodicity, symmetry, and back-to-back properties
shift
Image Enhancement in Frequency Domain 47
Additional properties of 2D Fourier
Transform
Separability
Image Enhancement in Frequency Domain 48
Implementation More on Periodicity
Convolution
f ( x ) h( x )
M 1
1
M
f ( m) h( x m)
m 0
Image Enhancement in Frequency Domain 49
extend
extend
Image Enhancement in Frequency Domain 50
Important Properties of 2-D Fourier
Transform
Image Enhancement in Frequency Domain 51
Important Properties of 2-D Fourier
Transform
Image Enhancement in Frequency Domain 52
Important Properties of 2-D Fourier
Transform
Image Enhancement in Frequency Domain 53
Important Properties of 2-D Fourier
Transform
Image Enhancement in Frequency Domain 54
Summary
We have looked at Fourier transform and its applications
in image enhancement:
Fourier transform and discrete Fourier transform for
image.
Ideal, Butterworth and Gaussian filters.
Image filtering in frequency domain.
Image Enhancement in Frequency Domain 55
