School of Info-Physics and Geomatics Engineering, CSU 1 2013/6/28

Chapter4 Image Enhancement in the

Frequency Domain
Enhance: To make greater (as in value, desirability, or
attractiveness.
Frequency: The number of times that a periodic
function repeats the same sequence of values during
a unit variation of the independent variable.
Websters New Collegiate Dictionary




School of Info-Physics and Geomatics Engineering, CSU 2 2013/6/28
Content
Background
Introduction to the Fourier Transform and the
Frequency Domain
Smoothing Frequency-Domain Filters
Sharpening Frequency-Domain Filters
Homomorphic Filtering
Implementation
Summary
This chapter is concerned primarily with helping the reader develop a
basic understanding of the Fourier transform and the frequency
domain, and how they apply to image enhancement.


School of Info-Physics and Geomatics Engineering, CSU 3 2013/6/28
4.1 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 advent of digital computation and the discovery of
fast Fourier Transform (FFT) algorithm in the late 1950s
revolutionized the field of signal processing, and allowed
for the first time practical processing and meaningful
interpretation of a host of signals of exceptional human
and industrial importance.


School of Info-Physics and Geomatics Engineering, CSU 4 2013/6/28
The frequency domain
refers to the plane of
the two dimensional
discrete Fourier
transform of an image.
The purpose of the
Fourier transform is to
represent a signal as a
linear combination of
sinusoidal signals of
various frequencies.


School of Info-Physics and Geomatics Engineering, CSU 5 2013/6/28
4.2 Introduction to the Fourier Transform
and the Frequency Domain
The one-dimensional Fourier transform and its inverse
Fourier transform (continuous case)

Inverse Fourier transform:

The two-dimensional Fourier transform and its inverse
Fourier transform (continuous case)


Inverse Fourier transform:

}

= du e u F x f
ux j t 2
) ( ) (
} }

+
= dy dx e y x f v u F
vy ux j ) ( 2
) , ( ) , (
t
1 where ) ( ) (
2
= =
}

j dx e x f u F
ux j t
} }

+
= dv du e v u F y x f
vy ux j ) ( 2
) , ( ) , (
t
u u
u
sin cos j e
j
+ =


School of Info-Physics and Geomatics Engineering, CSU 6 2013/6/28
4.2.1The one-dimensional Fourier transform and
its inverse (discrete time case)
Fourier transform (DFT)


Inverse Fourier transform (IDFT)



The 1/M multiplier in front of the Fourier transform sometimes
is placed in the front of the inverse instead. Other times both
equations are multiplied by
Unlike continuous case, the discrete Fourier transform and its
inverse always exist, only if f(x) is finite duration.


1 ,..., 2 , 1 , 0 for ) (
1
) (
1
0
/ 2
= =

M u e x f
M
u F
M
x
M ux j t
1 ,..., 2 , 1 , 0 for ) ( ) (
1
0
/ 2
= =

=
M x e u F x f
M
u
M ux j t
1/ M


School of Info-Physics and Geomatics Engineering, CSU 7 2013/6/28
Since and the fact
then discrete Fourier transform can be redefined



Frequency (time) domain: the domain (values of u) over
which the values of F(u) range; because u determines
the frequency of the components of the transform.
Frequency (time) component: each of the M terms of
F(u).

1
0
1
( ) ( )[cos 2 / sin 2 / ]
for 0,1, 2,..., 1
M
x
F u f x ux M j ux M
M
u M
t t

=
=
=

u u
u
sin cos j e
j
+ =
u u cos ) cos( =


School of Info-Physics and Geomatics Engineering, CSU 8 2013/6/28
F(u) can be expressed in polar coordinates:




R(u): the real part of F(u)
I(u): the imaginary part of F(u)
Power spectrum:
) ( ) ( ) ( ) (
2 2
2
u I u R u F u P + = =
( )
1/2
2 2
1
( ) ( )
where ( ) ( ) ( ) (magnitude or spectrum)
( )
( ) tan (phase angle or phase spectrum)
( )
j u
F u F u e
F u R u I u
I u
u
R u
|
|

=
( = +

(
=
(



School of Info-Physics and Geomatics Engineering, CSU 9 2013/6/28
Some One-Dimensional Fourier Transform Examples
Please note the relationship between the value of K and the height of
the spectrum and the number of zeros in the frequency domain.


School of Info-Physics and Geomatics Engineering, CSU 10 2013/6/28
The transform of a constant function is a DC value only.




The transform of a delta function is a constant.


School of Info-Physics and Geomatics Engineering, CSU 11 2013/6/28
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.


School of Info-Physics and Geomatics Engineering, CSU 12 2013/6/28
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.


School of Info-Physics and Geomatics Engineering, CSU 13 2013/6/28
4.2.2 The two-dimensional Fourier transform and
its inverse (discrete time case)
Fourier transform (DFT)




Inverse Fourier transform (IDFT)




u, v : the transform or frequency variables
x, y : the spatial or image variables
1 ,..., 2 , 1 , 0 , 1 ,..., 2 , 1 , 0 for
) , (
1
) , (
1
0
1
0
) / / ( 2
= =
=

=
+
N v M u
e y x f
MN
v u F
M
x
N
y
N vy M ux j t
1 ,..., 2 , 1 , 0 , 1 ,..., 2 , 1 , 0 for
) , ( ) , (
1
0
1
0
) / / ( 2
= =
=

=
+
N y M x
e v u F y x f
M
u
N
v
N vy M ux j t


School of Info-Physics and Geomatics Engineering, CSU 14 2013/6/28
We define the Fourier spectrum, phase anble, and power
spectrum of the two-dimensional Fourier transform as
follows:





R(u,v): the real part of F(u,v)
I(u,v): the imaginary part of F(u,v)
| |
spectrum) (power ) , ( ) , ( ) , ( ) (
angle) (phase
) , (
) , (
tan ) , (
spectrum) ( ) , ( ) , ( ) , (
2 2
2
1
2
1
2 2
v u I v u R v u F u,v P
v u R
v u I
v u
v u I v u R v u F
+ = =
(

=
+ =

|


School of Info-Physics and Geomatics Engineering, CSU 15 2013/6/28
Some properties of Fourier transform:





| |
) (symmetric ) , ( ) , (
symmetric) (conujgate ) , ( * ) , (
(average) ) , (
1
) 0 , 0 (
(shift) )
2
,
2
( ) 1 )( , (
1
0
1
0
v u F v u F
v u F v u F
y x f
MN
F
N
v
M
u F y x f
M
x
N
y
y x
=
=
=
=

=
+


School of Info-Physics and Geomatics Engineering, CSU 16 2013/6/28
(a) f(x,y) (b) F(u,y) (c) F(u,v)

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)

Steps and some example of two-dimensional DFT
y or v
x or u
Convention of coordination:


School of Info-Physics and Geomatics Engineering, CSU 17 2013/6/28
shift
Consider the relationship between the separation of zeros in u- or v- direction
and the width or height of white block of source image.


School of Info-Physics and Geomatics Engineering, CSU 18 2013/6/28
Shape of three dimensional spectrum


School of Info-Physics and Geomatics Engineering, CSU 19 2013/6/28
DFT
DFT
The Property of Two-Dimensional DFT Rotation


School of Info-Physics and Geomatics Engineering, CSU 20 2013/6/28
DFT
DFT
DFT
A
B
0.25 * A
+ 0.75 * B
The Property of Two-Dimensional DFT Linear Combination


School of Info-Physics and Geomatics Engineering, CSU 21 2013/6/28
DFT
DFT
A
Expanding the original image by a factor of n (n=2), filling
the empty new values with zeros, results in the same DFT.
B
The Property of Two-Dimensional DFT Expansion


School of Info-Physics and Geomatics Engineering, CSU 22 2013/6/28
Two-Dimensional DFT with Different Functions
Sine wave
Rectangle
Its DFT
Its DFT


School of Info-Physics and Geomatics Engineering, CSU 23 2013/6/28
Two-Dimensional DFT with Different Functions
2D Gaussian
function
Impulses
Its DFT
Its DFT


School of Info-Physics and Geomatics Engineering, CSU 24 2013/6/28
4.2.3 Filtering in the Frequency Domain
Frequency is directly related to rate of change. The frequency of fast
varying components in an image is higher than slowly varying components.


School of Info-Physics and Geomatics Engineering, CSU 25 2013/6/28
Basics of Filtering in the Frequency Domain
Including multiplication the input/output image by (-1)
x+y
.
What is zero-phase-shift filter?


School of Info-Physics and Geomatics Engineering, CSU 26 2013/6/28
Some Basic Filters and Their Functions
Multiply all values of F(u,v) by the filter function (notch filter):


All this filter would do is set F(0,0) to zero (force the average value of
an image to zero) and leave all other frequency components of the
Fourier transform untouched and make prominent edges stand out

=
=
otherwise. 1
) 2 / , 2 / ( ) , ( if 0
) , (
N M v u
v u H


School of Info-Physics and Geomatics Engineering, CSU 27 2013/6/28
Some Basic Filters and Their Functions
Lowpass filter
Highpass filter
Circular symmetry


School of Info-Physics and Geomatics Engineering, CSU 28 2013/6/28
Some Basic Filters and Their Functions
Low frequency filters: eliminate the gray-level detail and keep the general
gray-level appearance. (blurring the image)
Low frequency filters: have less gray-level variations in smooth areas and
emphasized transitional (e.g., edge and noise) gray-level detail. (sharpening
images)


School of Info-Physics and Geomatics Engineering, CSU 29 2013/6/28
4.2.4 Correspondence between Filtering in the
Spatial and Frequency Domain

Convolution theorem:
The discrete convolution of two functions f(x,y) and h(x,y) of
size M N is defined as




The process of implementation:
1) Flipping one function about the origin;
2) Shifting that function with respect to the other by changing the
values of (x, y);
3) Computing a sum of products over all values of m and n, for
each displacement.
1 1
0 0
1 1
0 0
1
( , ) ( , ) ( , ) ( , )
1
( , ) ( , )
M N
m n
M N
m n
f x y h x y f m n h x m y n
MN
h m n f x m y n
MN

= =

= =
- =
=

School of Info-Physics and Geomatics Engineering, CSU 30 2013/6/28
) , ( ) , ( ) , ( ) , ( v u H v u F y x h y x f -
) , ( ) , ( ) , ( ) , ( v u H v u F y x h y x f -
Eq. (4.2-
32)
Let F(u,v) and H(u,v) denote the Fourier transforms of f(x,y) and
h(x,y), then
Eq. (4.2-31)
an impulse function of strength A, located at coordinates
(x
0
,y
0
): and is defined by
:



where : a unit impulse located at the origin
The Fourier transform of a unit impulse at the origin (Eq4.2-35) :

=
=
1
0
1
0
0 0 0 0
) , ( ) , ( ) , (
M
x
N
y
y x As y y x x A y x s o
) , (
0 0
y y x x A o

=
=
1
0
1
0
) 0 , 0 ( ) , ( ) , (
M
x
N
y
s y x y x s o

=
+
= =
1
0
1
0
) / / ( 2
1
) , (
1
) , (
M
x
N
y
N vy M ux j
MN
e y x
MN
v u F
t
o
) , ( y x o
The shifting property of impulse
function


School of Info-Physics and Geomatics Engineering, CSU 31 2013/6/28
Let , then the convolution (Eq. (4.2-36))



Combine Eqs. (4.2-35) (4.2-36) with Eq. (4.2-31), we
obtain:


) , (
1

) , ( ) , (
1
) , ( ) , (
1
0
1
0
y x h
MN
n y m x h n m
MN
y x h y x f
M
m
N
n
=
= -

=
o
) , ( ) , ( y x y x f o =
| |
) , ( ) , (
) , (
1
) , (
1
) , ( ) , ( ) , ( ) , (
) , ( ) , ( ) , ( ) , (
v u H y x h
v u H
MN
y x h
MN
v u H y x y x h y x
v u H v u F y x h y x f

-
-
o o
That is to say, the response of
impulse input is the transfer
function of filter.


School of Info-Physics and Geomatics Engineering, CSU 32 2013/6/28
The distinction and links between spatial and frequency filtering
If the size of spatial and frequency filters is same, then the
computation burden in spatial domain is larger than in frequency
domain;
However, whenever possible, it makes more sense to filter in the
spatial domain using small filter masks.
Filtering in frequency is more intuitive. We can specify filters in the
frequency, take their inverse transform, and the use the resulting
filter in spatial domain as a guide for constructing smaller spatial
filter masks.
Fourier transform and its inverse are linear process, so the
following discussion is limited to linear processes.


School of Info-Physics and Geomatics Engineering, CSU 33 2013/6/28
Let H(u) denote a frequency domain, Gaussian filter
function given the equation

where : the standard deviation of the Gaussian curve.
The corresponding filter in the spatial domain is




2 2 2
2
2 ) (
x
Ae x h
o t
o t

=
2 2
2 /
) (
o u
Ae u H

=
o
There is two reasons that filters based on Gaussian functions are of
particular importance: 1) their shapes are easily specified; 2) both
the forward and inverse Fourier transforms of a Gaussian are real
Gaussian function.


School of Info-Physics and Geomatics Engineering, CSU 34 2013/6/28


School of Info-Physics and Geomatics Engineering, CSU 35 2013/6/28
2 2 2 2
1 2
/ 2 / 2
1 2
( ) , ( , )
u u
H u Ae Be A B
o o
o o

= > >
The corresponding filter in the spatial domain is
2 2 2 2 2 2
1 2
2 2
1 2
( ) 2 2
x x
h x Ae Be
t o t o
to to

=
We can note that the value of this types of filter has both negative and
positive values. Once the values turn negative, they never turn positive again.

Filtering in frequency domain is usually used for the guides to design the
filter masks in the spatial domain.


School of Info-Physics and Geomatics Engineering, CSU 36 2013/6/28
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 o
2
in the equations.
These properties make the Gaussian filter very useful for lowpass
filtering an image. The amount of blur is controlled by o
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.
Some important properties of Gaussian filters funtions


School of Info-Physics and Geomatics Engineering, CSU 37 2013/6/28
4.2 Smoothing Frequency-Domain Filters
The basic model for filtering in the frequency domain

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
) , ( ) , ( ) , ( v u F v u H v u G =


School of Info-Physics and Geomatics Engineering, CSU 38 2013/6/28
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 D
0

from the origin of the transform.
The transfer function of an ideal lowpass filter


where D(u,v) : the distance from point (u,v) to the center
of ther frequency rectangle (M/2, N/2)

| |
2
1
2 2
) 2 / ( ) 2 / ( ) , ( N v M u v u D + =

>
s
=
) , ( if 0
) , ( if 1
) , (
0
0
D v u D
D v u D
v u H


School of Info-Physics and Geomatics Engineering, CSU 39 2013/6/28
cutoff frequency
ILPF is a type of nonphysical filters and cant be realized with electronic
components and is not very practical.


School of Info-Physics and Geomatics Engineering, CSU 40 2013/6/28


School of Info-Physics and Geomatics Engineering, CSU 41 2013/6/28
The blurring and ringing
phenomena can be seen, in
which ringing behavior is
characteristic of ideal filters.


School of Info-Physics and Geomatics Engineering, CSU 42 2013/6/28
Another example of
ILPF
Figure 4.13 (a) A frequency-domain
ILPF of radius 5. (b) Corresponding
spatial filter. (c) Five impulses in the
spatial domain, simulating the values
of five pixels. (d) Convolution of (b)
and (c) in the spatial domain.
frequency
spatial
spatial
spatial
) , ( ) , ( ) , ( ) , ( v u H v u F y x h y x f -
Notation: the radius of center
component and the number of
circles per unit distance from
the origin are inversely
proportional to the value of the
cutoff frequency.
diagonal scan line of (d)


School of Info-Physics and Geomatics Engineering, CSU 43 2013/6/28
| |
n
D v u D
v u H
2
0
/ ) , ( 1
1
) , (
+
=
Butterworth Lowpass Filters (BLPFs) with order n
Note the relationship
between order n and
smoothing
The BLPF may be viewed as a transition between ILPF AND GLPF, BLPF of
order 2 is a good compromise between effective lowpass filtering and
acceptable ringing characteristics.


School of Info-Physics and Geomatics Engineering, CSU 44 2013/6/28
Butterworth Lowpass
Filters (BLPFs)


n=2
D
0
=5,15,30,80,and 230


School of Info-Physics and Geomatics Engineering, CSU 45 2013/6/28
Butterworth Lowpass Filters (BLPFs)
Spatial Representation
n=1
n=2 n=5
n=20


School of Info-Physics and Geomatics Engineering, CSU 46 2013/6/28
Gaussian Lowpass Filters (FLPFs)
2
0
2
2 / ) , (
) , (
D v u D
e v u H

=


School of Info-Physics and Geomatics Engineering, CSU 47 2013/6/28
Gaussian Lowpass
Filters (FLPFs)


D
0
=5,15,30,80,and 230


School of Info-Physics and Geomatics Engineering, CSU 48 2013/6/28
Additional Examples of Lowpass Filtering
Character recognition in machine perception: join the broken character
segments with a Gaussian lowpass filter with D
0
=80.


School of Info-Physics and Geomatics Engineering, CSU 49 2013/6/28
Application in cosmetic processing and produce a smoother,
softer-looking result from a sharp original.


School of Info-Physics and Geomatics Engineering, CSU 50 2013/6/28
Gaussian lowpass filter for reducing the horizontal sensor scan lines and
simplifying the detection of features like the interface boundaries.


School of Info-Physics and Geomatics Engineering, CSU 51 2013/6/28
( , ) 1 ( , )
hp lp
H u v H u v =
Ideal highpass filter
Butterworth highpass filter
Gaussian highpass filter

>
s
=
) , ( if 1
) , ( if 0
) , (
0
0
D v u D
D v u D
v u H
| |
n
v u D D
v u H
2
0
) , ( / 1
1
) , (
+
=
2
0
2
2 / ) , (
1 ) , (
D v u D
e v u H

=
4.4 Sharpening Frequency Domain Filter


School of Info-Physics and Geomatics Engineering, CSU 52 2013/6/28
Highpass Filters Spatial Representations


School of Info-Physics and Geomatics Engineering, CSU 53 2013/6/28
Ideal Highpass Filters (IHPFs)

>
s
=
) , ( if 1
) , ( if 0
) , (
0
0
D v u D
D v u D
v u H
non-physically realizable with electronic component and have the
same ringing properties as ILPFs.


School of Info-Physics and Geomatics Engineering, CSU 54 2013/6/28
Butterworth Highpass Filters
| |
n
v u D D
v u H
2
0
) , ( / 1
1
) , (
+
=
The result is smoother than that of IHPFs and sharper than that of GHPFs


School of Info-Physics and Geomatics Engineering, CSU 55 2013/6/28
Gaussian Highpass Filters
2
0
2
2 / ) , (
1 ) , (
D v u D
e v u H

=
The result is the smoothest in three types of high-pass filters


School of Info-Physics and Geomatics Engineering, CSU 56 2013/6/28
The Laplacian in the Frequency Domain
) ( ) , (
2 2
v u v u H + =
2 2
( , ) ( / 2) ( / 2) H u v u M v N ( = +

The FT of n-order differential of a function f(x) is
( ) / ( ) ( )
n n n
d f x dx ju F u ( =

2 2
2 2 2 2 2
2 2
( , ) ( , )
[ ( , )] [ ] ( ) ( , ) ( ) ( , ) ( ) ( , )
f x y f x y
f x y ju F u v jv F u v u v F u v
x y
c c
V = + = + = +
c c
For a two-dimensional function f(x,y), it can be shown that
So, Laplacian can be implemented in the frequency domain by using the filter
Shift the center to (M/2, N/2) and obtain
We have the following Fourier transform pairs
2 2 2
( , ) ( / 2) ( / 2) ( , ) f x y u M v N F u v ( V +



School of Info-Physics and Geomatics Engineering, CSU 57 2013/6/28
Frequency
domain
Spatial domain
The plot of Laplacian in frequency and spatial domain


School of Info-Physics and Geomatics Engineering, CSU 58 2013/6/28
2 1 2 2
( , ) ( , ) ( , ) 1 (( ) ( ) ) ( , )
2 2
M N
g x y f x y f x y u v F u v

(
= V = + +
`
(

)
For display
purposes only
A integrated operation
in frequency domain



School of Info-Physics and Geomatics Engineering, CSU 59 2013/6/28
Unsharp masking, high-boost filtering, and high-frequency emphasis
filtering (refers to page187-191)


School of Info-Physics and Geomatics Engineering, CSU 60 2013/6/28
4.5 Homomorphic filtering



Problems:
When the illumination radiating to an object is non-
uniform, the detail of the dark part in the image is
more discernable.
aims:
Simultaneously compress the gray-level range and
enhance contrast, eliminate the effect of non-uniform
illumination, and emphasis the details.
Principal:
Generally, the illumination component of an image is
characterized by slow spatial variations, while the
reflectance components tends to vary abruptly,
particularly at the junctions of dissimilar objects.
These characteristics lead to associating the low
frequencies of the Fourier transform of the logarithm
of an image with illumination and the high
frequencies with reflectance.



School of Info-Physics and Geomatics Engineering, CSU 61 2013/6/28
) , ( ) , ( ) , ( y x r y x i y x f =
The illumination-reflectance model of an image
) , ( y x i ) , ( y x r Illumination coefficient: reflectance coefficient:
Steps:
) , ( ln ) , ( ln ) , ( ln ) , ( y x r y x i y x f y x z + = =
)] , ( [ln )] , ( [ln )] , ( [ y x r y x i y x z F F F + =
) , ( ) , ( ) , ( v u R v u I v u Z + =
) , ( ) , ( ) , ( ) , ( ) , ( v u R v u H v u I v u H v u S + =
2)
1)
3)
Determine the H(u, v), which must compress the dynamic
range of i(x,y), and enhance the contrast of r(x,y) component.


School of Info-Physics and Geomatics Engineering, CSU 62 2013/6/28
r
L
<1
r
H
>1
2 2
0
( ( , )/ )
( , ) ( )[1 ]
c D u v D
H L L
H u v e

= +
The following function meet the above requires
The curve shape shown in above figure can be approximated using basic form
of the ideal highpass filters, for example, using a slightly modified form of the
Gaussian highpass filter and can obtain


School of Info-Physics and Geomatics Engineering, CSU 63 2013/6/28
)] , ( ) , ( [ ) , (
1 '
v u I v u H y x i

=F
)] , ( ) , ( [ ) , (
1 '
v u R v u H y x r

=F
)] , ( exp[ ) , (
'
0
y x i y x i =
)] , ( exp[ ) , (
'
0
y x r y x r =
) , ( ) , ( ) , (
0 0
y x r y x i y x g =
Steps: 4)
5)


School of Info-Physics and Geomatics Engineering, CSU 64 2013/6/28
ln FFT H(u,v)
FFT
-1

exp
f(x,y)
g(x,y)
The flow-chart of Homomorphic filtering


School of Info-Physics and Geomatics Engineering, CSU 65 2013/6/28
Two examples


School of Info-Physics and Geomatics Engineering, CSU 66 2013/6/28


School of Info-Physics and Geomatics Engineering, CSU 67 2013/6/28
4.6 Implementation
4.6.1Some Additional Properties of the 2D Fourier Transform
, distributivity, scaling, and :
0 0
0 0
2 ( / / )
0 0
2 ( / / )
0 0
( , ) ( , )
( , ) ( , )
j ux M vy N
j xu M yu N
f x x y y F u v e
f x y e F u u v v
t
t
+
+


1 2 1 2
[ ( , ) ( , )] [ ( , )] [ ( , )]
1
( , ) ( / , / )
f x y f x y f x y f x y
f ax by F u a v b
ab
+ = +

0 0
( , ) ( , ) f r F w u u u + +
Translation
Distributivity and
scaling
rotation


School of Info-Physics and Geomatics Engineering, CSU 68 2013/6/28
Periodicity, conjugate symmetry, and back-to-back properties
shift
( , ) ( , ) ( , ) ( , )
( , ) ( , ) ( , ) ( , )
( , ) ( , )
F u v F u M v F u v N F u M v N
f u v f u M v f u v N f u M v N
F u v F u v
-
= + = + = + +
= + = + = + +
=


School of Info-Physics and Geomatics Engineering, CSU 69 2013/6/28
Separability
1 1 1
2 / 2 / 2 /
0 0 0
1 1 1
( , ) ( , ) ( , )
M N M
j ux M j vy N j ux M
x y x
F u v e f x y e F x v e
M N M
t t t


= = =
= =

A similar process can be applied to computing the 2-D inverse Fourier
transform.


School of Info-Physics and Geomatics Engineering, CSU 70 2013/6/28
4.6.2 computing the inverse FF using a forward
transform algorithm
Repeat the one-dimensional inverse FF:
1
2 /
0
( ) ( )
M
j ux M
u
f x F u e
t

=
=

Take the complex conjugate of two side and multiply M

1
2 /
0
1 1
( ) ( )
M
j ux M
u
f x F u e
M M
t

- -
=
=

Which is the form of forward FF. Take the complex conjugate of the result of
the above equation and will get the inverse FF by forward transform.
For two-dimensional case, similarly have
1 1
2 ( / / )
0 0
1 1
( , ) ( , )
M N
j ux M vy N
u v
f x y F u v e
MN MN
t

- - +
= =
=

Here, we can treat F(u,v) as a simple function presenting on the forward
transform equation


School of Info-Physics and Geomatics Engineering, CSU 71 2013/6/28
4.6.3 More on Periodicity: the need for padding
1
0
( ) ( )
1
( ) ( )
M
m
f x h x
f m h x m
M

=
- =

Convolution process
Aliasing or wraparound error


School of Info-Physics and Geomatics Engineering, CSU 72 2013/6/28
extend
extend
The methods of solving the
aliasing problem is to
extend and pad.
( ) 0 1
( )
0 1
( ) 0 1
( )
0 1
where 1
e
e
f x x A
f x
A x P
h x x B
h x
B x P
P A+B
s s

s s

s s

s s

>


School of Info-Physics and Geomatics Engineering, CSU 73 2013/6/28
For two-dimensional case


School of Info-Physics and Geomatics Engineering, CSU 74 2013/6/28
An example
*
=


School of Info-Physics and Geomatics Engineering, CSU 75 2013/6/28
4.6.4 convolution and correlation theorems
1 1
0 0
1
( , ) ( , ) ( , ) ( , )
M N
m n
f x y h x y f m n h x m y n
MN

= =
- =

1 1
0 0
1
( , ) ( , ) ( , ) ( , )
M N
m n
f x y h x y f m n h x m y n
MN

-
= =
= + +

Except for the complex of f and h not mirrored about the origin, everything
else in the implementation of correlation is identical to convolution,
including the need for padding.
convolution
correlation
( , ) ( , ) ( , ) ( , )
( , ) ( , ) ( , ) ( , )
f x y h x y F u v H u v
f x y h x y F u v H u v
-
-
=
=
Correlation theorem
Correlation includes across- and auto-correlation, and its main use is for
matching and sure the location where h (template) finds a correspondence
in f.


School of Info-Physics and Geomatics Engineering, CSU 76 2013/6/28


School of Info-Physics and Geomatics Engineering, CSU 77 2013/6/28
4.6.5 Summary of Some Important Properties
of the 2-D Fourier Transform


School of Info-Physics and Geomatics Engineering, CSU 78 2013/6/28


School of Info-Physics and Geomatics Engineering, CSU 79 2013/6/28


School of Info-Physics and Geomatics Engineering, CSU 80 2013/6/28


School of Info-Physics and Geomatics Engineering, CSU 81 2013/6/28
Summary
The basic concept about Fourier Transform (FT)
The algorithm of FT and the process of frequency filtering
The physical meaning related to FT
The relationship of resolution between spatial and frequency
domain
Correspondence between filtering in the spatial and frequency
domains
The general type of smoothing and sharpening filters and their
main features
Homomorphic filtering
Some important properties of FT
Convolution and correlation theorems
Spatial and frequency filtering are both highly subjective processes