Sunteți pe pagina 1din 19

DIP Lecture

Fourier Transform

Fourier Transform: Concept


The Fourier transform is a mathematical
transformation employed to transform signals
between time (or spatial) domain and
frequency domain.

The motivation for the Fourier transform


comes from the study of Fourier series. In the
study of Fourier series, complicated but

periodic functions are written as the


weighted sum of simple waves
mathematically represented by sines and
cosines.
Fourier Transform is a change of basis, where
the basis functions consist of sines and cosines
(complex exponentials).
2

Applications of Fourier transforms


1-D Fourier transforms are used in Signal Processing
2-D Fourier transforms are used in Image Processing
Applications of Fourier transforms in Image processing:

Image enhancement,
Image restoration,
Image encoding / decoding,
Image description

Motivation
1. Often interested in the frequency
content of signals (including images)
E.g., noise may have different
frequency content than signal of
interest.

2. Fourier Transforms makes analysis


and design of filters more intuitive than
when using convolution since we are
often interested in manipulating the
frequency content of a signal.

We can manipulate Fourier transform


(represents frequency content) of a
signal and then take inverse transform.

Motivation . Compression

3/19/2016

Muhammad Ali Jinnah University Islamabad


Advanced Digital Image Processing EE5623

Motivation
D0 5 pixels

D0 15 pixels

D0 30 pixels

D0 80 pixels

D0 230 pixels
Muhammad Ali Jinnah University Islamabad
Advanced Digital Image Processing EE5623

Periodic Square Wave

Can square wave be reproduced with cosine waves?

Relation between Periodic signals in


time Vs Fourier Series representation

Approximation of periodic Square


Wave Vs Fourier Series
Representation

Math Review - Periodic Functions


A periodic function is an oscillating (wave-like)function which repeats a pattern (values)
at regular intervals.
One complete repetition of the pattern is called a cycle. The period of a function is the
horizontal length of one complete cycle.

If there is some a, for a function f(x), such that


f(x) = f(x + na)
then function is periodic with the period a

Math Review - Attributes of cosine wave


http://www.regentsprep.org/regents/math/alg
trig/att7/graphvocab.htm

5
3
1

f(x) = cos (x)

-10

-5

-1 0

10

10

-3
-5

Amplitude

4
3
2
1

f(x) = 5 cos (x)

0
-10

-5

-1 0
-2
-3
-4
-5

Phase:
(initial angle of a
sinusoidal function at its
origin)

5
3

f(x) = 5 cos (x + )

1
-10

-5

-1 0

10

10

-3

Here =pi=3.14

-5

Frequency:
(number of cycles
completed in a given
interval)

f(x) = 5 cos (3 x + )

3
1
-10

-5

-1 0
-3
-5

Math Review - Attributes of cosine


wave
5
3
1
-10

-5

-1 0
-3
-5

f(x) = A cos (kx + )


Amplitude, Frequency, Phase

10

Math Review - Complex numbers


Amplitude
General Form
Z = a + bi
Re(Z) = a
Im(Z) = b

A = | Z | = (a2 + b2)

Phase
= Z = tan-1(b/a)

Math Review Complex Numbers


Polar Coordinate
Z = a + bi
Z=A(cos + i sin )

Eulers formula
a

Amplitude

A = (a2 + b2)

Phase

= tan-1(b/a)

Fourier Transform: 1D Cont. Signals


Fourier Transform of a 1D continuous signal

F (u )

f ( x)e j 2 ux dx

Eulers formula

e j 2 ux cos 2 ux j sin 2 ux

Inverse Fourier Transform

f ( x)

F (u )e j 2 ux du

17

Discrete Fourier Transforms (DFT)


1-D DFT for M samples is given as

1
F (u )
M

M 1

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

x 0

e j cos j sin ,

In Matlab
F=fft(f)

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

M x 0
F (u ) R (u ) jI (u )
Real

Imaginary

The inverse Fourier transform in 1-D is given as


M 1

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


u 0

In Matlab
f=ifft(F)

1-D Discrete Fourier Transform Example

S-ar putea să vă placă și