![Page 1: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/1.jpg)
Fourier Transform
1
![Page 2: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/2.jpg)
The idea A signal can be interpreted as en electromagnetic wave. This consists of lights of different “color”, or frequency, that can be split apart usign an optic prism. Each component is a “monochromatic” light with sinusoidal shape.
Following this analogy, each signal can be decomposed into its “sinusoidal” components which represent its “colors”.
Of course these components in general do not correspond to visible monochromatic light. However, they give an idea of how fast are the changes of the signal.
2
![Page 3: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/3.jpg)
Contents
• Signals as functions (1D, 2D) – Tools
• Continuous Time Fourier Transform (CTFT)
• Discrete Time Fourier Transform (DTFT)
• Discrete Fourier Transform (DFT)
• Discrete Cosine Transform (DCT)
• Sampling theorem
3
![Page 4: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/4.jpg)
Fourier Transform
• Different formulations for the different classes of signals – Summary table: Fourier transforms with various combinations of continuous/
discrete time and frequency variables. – Notations:
• CTFT: continuous time FT: t is real and f real (f=ω) (CT, CF) • DTFT: Discrete Time FT: t is discrete (t=n), f is real (f=ω) (DT, CF) • CTFS: CT Fourier Series (summation synthesis): t is real AND the function is periodic, f
is discrete (f=k), (CT, DF) • DTFS: DT Fourier Series (summation synthesis): t=n AND the function is periodic, f
discrete (f=k), (DT, DF) • P: periodical signals • T: sampling period • ωs: sampling frequency (ωs=2π/T) • For DTFT: T=1 → ωs=2π
• This is a hint for those who are interested in a more exhaustive theoretical approach
4
![Page 5: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/5.jpg)
Images as functions
• Gray scale images: 2D functions – Domain of the functions: set of (x,y) values for which f(x,y) is defined : 2D lattice
[i,j] defining the pixel locations – Set of values taken by the function : gray levels
• Digital images can be seen as functions defined over a discrete domain {i,j: 0<i<I, 0<j<J}
– I,J: number of rows (columns) of the matrix corresponding to the image – f=f[i,j]: gray level in position [i,j]
5
![Page 6: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/6.jpg)
Mathematical Background: Complex Numbers
• A complex number x is of the form:
a: real part, b: imaginary part
• Addition
• Multiplication
6
![Page 7: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/7.jpg)
Mathematical Background: Complex Numbers (cont’d)
• Magnitude-Phase (i.e.,vector) representation
Magnitude:
Phase:
Phase – Magnitude notation:
7
![Page 8: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/8.jpg)
Mathematical Background: Complex Numbers (cont’d)
• Multiplication using magnitude-phase representation
• Complex conjugate
• Properties
8
![Page 9: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/9.jpg)
Mathematical Background: Complex Numbers (cont’d)
• Euler’s formula
• Properties
9
![Page 10: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/10.jpg)
Mathematical Background: Sine and Cosine Functions
• Periodic functions
• General form of sine and cosine functions:
10
![Page 11: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/11.jpg)
Mathematical Background: Sine and Cosine Functions
Special case: A=1, b=0, a=1
π
π
11
![Page 12: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/12.jpg)
Mathematical Background: Sine and Cosine Functions (cont’d)
Note: cosine is a shifted sine function:
• Shifting or translating the sine function by a const b
cos( ) sin( )2
t t π= +
12
![Page 13: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/13.jpg)
Mathematical Background: Sine and Cosine Functions (cont’d)
• Changing the amplitude A
13
![Page 14: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/14.jpg)
Mathematical Background: Sine and Cosine Functions (cont’d)
• Changing the period T=2π/|α| consider A=1, b=0: y=cos(αt)
period 2π/4=π/2
shorter period higher frequency (i.e., oscillates faster)
α =4
Frequency is defined as f=1/T
Alternative notation: sin(αt)=sin(2πt/T)=sin(2πft)
14
![Page 15: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/15.jpg)
Fourier Series Theorem
• Any periodic function can be expressed as a weighted sum (infinite) of sine and cosine functions of varying frequency:
is called the “fundamental frequency”
15
![Page 16: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/16.jpg)
Fourier Series (cont’d)
α1
α2
α3
16
![Page 17: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/17.jpg)
Concept
17
![Page 18: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/18.jpg)
Continuous Time Fourier Transform (FT)
• Transforms a signal (i.e., function) from the spatial domain to the frequency domain.
where
18
F ω( ) = f t( )−∞
+∞
∫ e− jωtdt
f t( ) = F ω( )−∞
+∞
∫ e jωtdω
Time domain
Spatial domain
![Page 19: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/19.jpg)
CTFT
• Change of variables for simplified notations: ω=2πu
• More compact notations (same as in GW)
2(2 ) ( ) ( ) j uxF u F u f x e dxππ∞
−
−∞
= = =∫( )2 21( ) ( ) 2 ( )
2j ux j uxf x F u e d u F u e duπ ππ
π
∞ ∞
−∞ −∞
= =∫ ∫
( )
2
2
( ) ( )
( )
j ux
j ux
F u f x e dx
f x F u e du
π
π
∞−
−∞
∞
−∞
=
=
∫
∫
19
![Page 20: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/20.jpg)
Images vs Signals
1D
• Signals
• Frequency – Temporal – Spatial
• Time (space) frequency characterization of signals
• Reference space for – Filtering – Changing the sampling rate – Signal analysis – ….
2D
• Images
• Frequency – Spatial
• Space/frequency characterization of 2D signals
• Reference space for – Filtering – Up/Down sampling – Image analysis – Feature extraction – Compression – ….
20
![Page 21: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/21.jpg)
Example: Removing undesirable frequencies
remove high frequencies
reconstructed signal
frequencies noisy signal
To remove certain frequencies, set their corresponding F(u) coefficients to zero!
21
![Page 22: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/22.jpg)
Frequency Filtering Steps
1. Take the FT of f(x):
2. Remove undesired frequencies:
3. Convert back to a signal:
We’ll talk more about this later .....
22
![Page 23: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/23.jpg)
Definitions
• F(u) is a complex function:
• Magnitude of FT (spectrum):
• Phase of FT:
• Magnitude-Phase representation:
• Energy of f(x): P(u)=|F(u)|2
23
![Page 24: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/24.jpg)
Continuous Time Fourier Transform (CTFT)
Time is a real variable (t)
Frequency is a real variable (ω)
Signals : 1D
24
![Page 25: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/25.jpg)
CTFT: Concept
■ A signal can be represented as a weighted sum of sinusoids.
■ Fourier Transform is a change of basis, where the basis functions consist of sins and cosines (complex exponentials).
[Gonzalez Chapter 4]
25
![Page 26: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/26.jpg)
Continuous Time Fourier Transform (CTFT)
T=1
26
![Page 27: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/27.jpg)
Fourier Transform
• Cosine/sine signals are easy to define and interpret.
• Analysis and manipulation of sinusoidal signals is greatly simplified by dealing with related signals called complex exponential signals.
• A complex number has real and imaginary parts: z = x+j y
• The Eulero formula links complex exponential signals and trigonometric functions
( )e cos sinjr r jα α α= +cosα = e
iα + e−iα
2
sinα = eiα − e−iα
2i27
![Page 28: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/28.jpg)
CTFT
• Continuous Time Fourier Transform
• Continuous time a-periodic signal
• Both time (space) and frequency are continuous variables – NON normalized frequency ω is used
• Fourier integral can be regarded as a Fourier series with fundamental frequency approaching zero
• Fourier spectra are continuous – A signal is represented as a sum of sinusoids (or exponentials) of all
frequencies over a continuous frequency interval
( ) ( )
1( ) ( )2
j t
tj t
F f t e dt
f t F e d
ω
ω
ω
ω
ω ωπ
−=
=
∫
∫
analysis
synthesis
Fourier integral
28
![Page 29: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/29.jpg)
CTFT of real signals
• Real signals: each signal sample is a real number
• Property: the CTFT is Hermttian-symmetric -> the spectrum is symmetric
29
f t( )→ f̂ ω( )f −t( )→ f̂ −ω( ) = f̂ ∗ ω( )Proof
ℑ f −t( ){ }= f −t( )e− jωt dt =−∞
+∞
∫ f t '( )e jωt ' dt ' =−∞
+∞
∫ f̂ ∗ −ω( )
f̂ −ω( ) = f̂ ∗ ω( )
ω
![Page 30: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/30.jpg)
Sinusoids
• Frequency domain characterization of signals
Frequency domain (spectrum, absolute value of the transform)
Signal domain
( ) ( ) j tF f t e dtωω+∞
−
−∞
= ∫
30
![Page 31: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/31.jpg)
Gaussian
Frequency domain
Time domain
31
![Page 32: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/32.jpg)
rect
sinc function
Frequency domain
Time domain
32
![Page 33: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/33.jpg)
Example
33
![Page 34: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/34.jpg)
Properties qui
34
![Page 35: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/35.jpg)
Discrete Fourier Transform (DFT)
Applies to finite length discrete time (sampled) signals and time series
The easiest way to get to it
Time is a discrete variable (t=n)
Frequency is a discrete variable (f=k)
35
![Page 36: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/36.jpg)
DFT
• The DFT can be considered as a generalization of the CTFT to discrete series
• It is the FT of a discrete (sampled) function of one variable
– The 1/N factor is put either in the analysis formula or in the synthesis one, or the 1/sqrt(N) is put in front of both.
• Calculating the DFT takes about N2 calculations
12 /
01
2 /
0
1[ ] [ ]
[ ] [ ]
Nj kn N
nN
j kn N
k
F k f n eN
f n F k e
π
π
−−
=−
=
=
=
∑
∑
36
![Page 37: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/37.jpg)
In practice..
• In order to calculate the DFT we start with k=0, calculate F(0) as in the formula below, then we change to u=1 etc
• F[0] is the mean value of the function f[n] – This is also the case for the CTFT
• The transformed function F[k] has the same number of terms as f[n] and always exists
• The transform is always reversible by construction so that we can always recover f given F
1 12 0 /
0 0
1 1[0] [ ] [ ]N N
j n N
n nF f n e f n f
N Nπ
− −−
= =
= = =∑ ∑
37
![Page 38: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/38.jpg)
Highlights on DFT properties
time
amplitude
0
Frequency (k)
|F[k]| F[0] low-pass
characteristic
38
The DFT of a real signal is symmetric (Hermitian symmetry) The DFT of a real symmetric signal (even like the cosine) is real and symmetric The DFT is N-periodic Hence The DFT of a real symmetric signal only needs to be specified in [0, N/2]
0 N/2 N
![Page 39: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/39.jpg)
Visualization of the basic repetition
• To show a full period, we need to translate the origin of the transform at u=N/2 (or at (N/2,N/2) in 2D)
|F(u-N/2)|
|F(u)|
f n[ ]e2πu0n → f k −u0[ ]
u0 =N2
f n[ ]e2π N
2n= f n[ ]eπNn = −1( )n f n[ ]→ f k − N
2#
$%&
'(
39
![Page 40: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/40.jpg)
DFT
• About N2 multiplications are needed to calculate the DFT
• The transform F[k] has the same number of components of f[n], that is N
• The DFT always exists for signals that do not go to infinity at any point
• Using the Eulero’s formula
( ) ( )( )1 1
2 /
0 0
1 1[ ] [ ] [ ] cos 2 / sin 2 /N N
j kn N
n nF k f n e f n j kn N j j kn N
N Nπ π π
− −−
= =
= = −∑ ∑
frequency component k discrete trigonometric functions
40
![Page 41: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/41.jpg)
2D example
no translation after translation
41
![Page 42: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/42.jpg)
Going back to the intuition
• The FT decomposed the signal over its harmonic components and thus represents it as a sum of linearly independent complex exponential functions
• Thus, it can be interpreted as a “mathematical prism”
42
![Page 43: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/43.jpg)
DFT
• Each term of the DFT, namely each value of F[k], results of the contributions of all the samples in the signal (f[n] for n=1,..,N)
• The samples of f[n] are multiplied by trigonometric functions of different frequencies
• The domain over which F[k] lives is called frequency domain
• Each term of the summation which gives F[k] is called frequency component of harmonic component
43
![Page 44: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/44.jpg)
DFT is a complex number
• F[k] in general are complex numbers
[ ] [ ]{ } [ ]{ }[ ] [ ] [ ]{ }
[ ] [ ]{ } [ ]{ }
[ ][ ]{ }[ ]{ }
[ ] [ ]
2 2
1
2
Re Im
exp
Re Im
Imtan
Re
F k F k j F k
F k F k j F k
F k F k F k
F kF k
F k
P k F k
−
= +
=
⎧ ⎫= +⎪ ⎪⎪ ⎪⎨ ⎬⎧ ⎫
= −⎪ ⎪⎨ ⎬⎪ ⎪⎩ ⎭⎩ ⎭
=
R
R
magnitude or spectrum
phase or angle
power spectrum
44
![Page 45: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/45.jpg)
Stretching vs shrinking
45
stretched shrinked
![Page 46: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/46.jpg)
Periodization vs discretization
• DT (discrete time) signals can be seen as sampled versions of CT (continuous time) signals
• Both CT and DT signals can be of finite duration or periodic
• There is a duality between periodicity and discretization – Periodic signals have discrete frequency (sampled) transform – Discrete time signals have periodic transform – DT periodic signals have discrete (sampled) periodic transforms
t
ampl
itude
ampl
itude
n
fk=f(kTs)
Ts
46
Linking continuous and discrete domains
![Page 47: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/47.jpg)
Increasing the resolution by Zero Padding
• Consider the analysis formula
• If f[n] consists of N samples than F[k] consists of N samples as well, it is discrete (k is an integer) and it is periodic (because the signal f[n] is discrete time, namely n is an integer)
• The value of each F[k], for all k, is given by a weighted sum of the values of f[n], for n=1,.., N-1
• Key point: if we artificially increase the length of the signal adding M zeros on the right, we get a signal f1[m] for which m=1,…,N+M-1. Since
47
F k[ ] = 1N
f n[ ]e−2π jknN
n=0
N−1
∑
f1 m[ ] =f [m] for 0 ≤m < N
0 for N ≤m < N +M
"
#$
%$
![Page 48: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/48.jpg)
Increasing the resolution through ZP
• Then the value of each F[k] is obtained by a weighted sum of the “real” values of f[n] for 0≤k≤N-1, which are the only ones different from zero, but they happen at different “normalized frequencies” since the frequency axis has been rescaled. In consequence, F[k] is more “densely sampled” and thus features a higher resolution.
48
![Page 49: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/49.jpg)
Increasing the resolution by Zero Padding
n
zero padding
0 N0
k 0 2π
F(Ω) (DTFT) in shade F[k]: “sampled version”
4π 2π/N0
49
![Page 50: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/50.jpg)
Zero padding
n
zero padding
0 N0
k 0 2π
F(Ω)
4π 2π/N0
Increasing the number of zeros augments the “resolution” of the transform since the samples of the DFT get “closer”
50
![Page 51: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/51.jpg)
Summary of dualities
FOURIER DOMAIN SIGNAL DOMAIN
Sampling Periodicity
Sampling Periodicity
DTFT
CTFS
Sampling+Periodicity Sampling +Periodicity DTFS/DFT
51
![Page 52: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/52.jpg)
Discrete Cosine Transform (DCT)
Applies to digital (sampled) finite length signals AND uses only cosines.
The DCT coefficients are all real numbers
52
![Page 53: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/53.jpg)
Discrete Cosine Transform (DCT)
• Operate on finite discrete sequences (as DFT)
• A discrete cosine transform (DCT) expresses a sequence of finitely many data points in terms of a sum of cosine functions oscillating at different frequencies
• DCT is a Fourier-related transform similar to the DFT but using only real numbers
• DCT is equivalent to DFT of roughly twice the length, operating on real data with even symmetry (since the Fourier transform of a real and even function is real and even), where in some variants the input and/or output data are shifted by half a sample
• There are eight standard DCT variants, out of which four are common
• Strong connection with the Karunen-Loeven transform – VERY important for signal compression
53
![Page 54: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/54.jpg)
DCT
• DCT implies different boundary conditions than the DFT or other related transforms
• A DCT, like a cosine transform, implies an even periodic extension of the original function
• Tricky part – First, one has to specify whether the function is even or odd at both the left and
right boundaries of the domain – Second, one has to specify around what point the function is even or odd
• In particular, consider a sequence abcd of four equally spaced data points, and say that we specify an even left boundary. There are two sensible possibilities: either the data is even about the sample a, in which case the even extension is dcbabcd, or the data is even about the point halfway between a and the previous point, in which case the even extension is dcbaabcd (a is repeated).
54
![Page 55: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/55.jpg)
Symmetries
55
![Page 56: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/56.jpg)
DCT
0
0
1
00 0
1
000 0
1cos 0,...., 12
2 1 1cos2 2
N
k nn
N
n kk
X x n k k NN
kx X X kN N
π
π
−
=
−
=
⎡ ⎤⎛ ⎞= + = −⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦
⎧ ⎫⎡ ⎤⎛ ⎞= + +⎨ ⎬⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦⎩ ⎭
∑
∑
• Warning: the normalization factor in front of these transform definitions is merely a convention and differs between treatments.
– Some authors multiply the transforms by (2/N0)1/2 so that the inverse does not require any additional multiplicative factor.
• Combined with appropriate factors of √2 (see above), this can be used to make the transform matrix orthogonal.
56
![Page 57: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/57.jpg)
Sampling
From continuous to discrete time
57
![Page 58: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/58.jpg)
58
Impulse Train
■ Define a comb function (impulse train) as follows
combN [n]= δ[n− lN ]l=−∞
∞
∑
where M and N are integers
2[ ]comb n
n
1
![Page 59: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/59.jpg)
59
Impulse Train
2[ ]comb n
n u
1 12
12
1 ( )2comb u
12
![Page 60: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/60.jpg)
60
Impulse Train
2( )comb x
x u
1 12
12
1 ( )2comb u
12
2
![Page 61: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/61.jpg)
Consequences
Sampling (Nyquist) theorem
61
![Page 62: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/62.jpg)
62
Sampling
x
xM
( )f x
( )Mcomb x
u
( )F u
u
1( )* ( )M
F u comb u
u1M
1 ( )M
comb u
x
( ) ( )Mf x comb x
![Page 63: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/63.jpg)
63
Sampling
x
( )f x
u
( )F u
u
1( )* ( )M
F u comb u
x
( ) ( )Mf x comb x
WW−
M
W
1M1 2W
M>Nyquist theorem: No aliasing if
![Page 64: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/64.jpg)
64
Sampling
u
1( )* ( )M
F u comb u
x
( ) ( )Mf x comb x
M
W
1M
If there is no aliasing, the original signal can be recovered from its samples by low-pass filtering.
12M
![Page 65: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/65.jpg)
65
Sampling
x
( )f x
u
( )F u
u
1( )* ( )M
F u comb u
( ) ( )Mf x comb x
WW−
W
1MAliased
![Page 66: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/66.jpg)
66
Sampling
x
( )f x
u
( )F u
u[ ]( )* ( ) ( )Mf x h x comb x
WW−
1M
Anti-aliasing filter
uWW−
( )* ( )f x h x12M
![Page 67: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/67.jpg)
67
Sampling
u[ ]( )* ( ) ( )Mf x h x comb x
1M
u( ) ( )Mf x comb x
W
1M
■ Without anti-aliasing filter:
■ With anti-aliasing filter:
![Page 68: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/68.jpg)
2D Continuous FT
68
![Page 69: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/69.jpg)
How do frequencies show up in an image?
• Low frequencies correspond to slowly varying information (e.g., continuous surface).
• High frequencies correspond to quickly varying information (e.g., edges)
Original Image Low-passed
69
![Page 70: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/70.jpg)
2D Frequency domain
ωx
ωy
Large vertical frequencies correspond to horizontal lines
Large horizontal frequencies correspond to vertical lines
Small horizontal and vertical frequencies correspond smooth grayscale changes in both directions
Large horizontal and vertical frequencies correspond sharp grayscale changes in both directions
70
![Page 71: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/71.jpg)
2D spatial frequencies
• 2D spatial frequencies characterize the image spatial changes in the horizontal (x) and vertical (y) directions
– Smooth variations -> low frequencies – Sharp variations -> high frequencies
x
y
ωx=1 ωy=0 ωx=0
ωy=1
71
![Page 72: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/72.jpg)
2D Continuous Fourier Transform
• 2D Continuous Fourier Transform (notation 2)
( ) ( ) ( )
( ) ( ) ( )
2
2
ˆ , ,
ˆ, ,
j ux vy
j ux vy
f u v f x y e dxdy
f x y f u v e dudv
π
π
+∞− +
−∞
+∞+
−∞
=
= =
∫
∫
22 ˆ( , ) ( , )f x y dxdy f u v dudv∞ ∞ ∞ ∞
−∞ −∞ −∞ −∞
=∫ ∫ ∫ ∫ Plancherel’s equality
72
![Page 73: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/73.jpg)
Delta
• Sampling property of the 2D-delta function (Dirac’s delta)
• Transform of the delta function
0 0 0 0( , ) ( , ) ( , )x x y y f x y dxdy f x yδ∞
−∞
− − =∫
{ } 2 ( )( , ) ( , ) 1j ux vyF x y x y e dxdyπδ δ∞ ∞
− +
−∞ −∞
= =∫ ∫
{ } 0 02 ( )2 ( )0 0 0 0( , ) ( , ) j ux vyj ux vyF x x y y x x y y e dxdy e ππδ δ
∞ ∞− +− +
−∞ −∞
− − = − − =∫ ∫ shifting property
73
![Page 74: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/74.jpg)
Constant functions
• Inverse transform of the impulse function
• Fourier Transform of the constant (=1 for all x and y)
( ) 2 ( )
( , ) 1 ,
, ( , )j ux vy
k x y x y
F u v e dxdy u vπ δ∞ ∞
− +
−∞ −∞
= ∀
= =∫ ∫
{ }1 2 ( ) 2 (0 0)( , ) ( , ) 1j ux vy j x vF u v u v e dudv eπ πδ δ∞ ∞
− + +
−∞ −∞
= = =∫ ∫
74
![Page 75: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/75.jpg)
Trigonometric functions
• Cosine function oscillating along the x axis – Constant along the y axis
{ } 2 ( )
2 ( ) 2 ( )2 ( )
( , ) cos(2 )
cos(2 ) cos(2 )
2
j ux vy
j fx j fxj ux vy
s x y fx
F fx fx e dxdy
e e e dxdy
π
π ππ
π
π π∞ ∞
− +
−∞ −∞
∞ ∞ −− +
−∞ −∞
=
= =
⎡ ⎤+= ⎢ ⎥
⎣ ⎦
∫ ∫
∫ ∫
[ ]
2 ( ) 2 ( ) 2
2 2 ( ) 2 ( ) 2 ( ) 2 ( )
12
1 112 2
1 ( ) ( )2
j u f x j u f x j vy
j vy j u f x j u f x j u f x j u f x
e e e dxdy
e dy e e dx e e dx
u f u f
π π π
π π π π π
δ δ
∞ ∞− − − + −
−∞ −∞
∞ ∞ ∞− − − − + − − − +
−∞ −∞ −∞
⎡ ⎤= + =⎣ ⎦
⎡ ⎤ ⎡ ⎤= + = + =⎣ ⎦ ⎣ ⎦
− + +
∫ ∫
∫ ∫ ∫
75
![Page 76: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/76.jpg)
Vertical grating
ωx
ωy
0
76
![Page 77: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/77.jpg)
Double grating
ωx
ωy
0
77
![Page 78: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/78.jpg)
Smooth rings
ωx
ωy
78
![Page 79: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/79.jpg)
Vertical grating
ωx
ωy
0 -2πf 2πf
79
![Page 80: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/80.jpg)
2D box 2D sinc
80
![Page 81: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/81.jpg)
CTFT properties
■ Linearity
■ Shifting
■ Modulation
■ Convolution
■ Multiplication
■ Separability
( , ) ( , ) ( , ) ( , )af x y bg x y aF u v bG u v+ ⇔ +
( , )* ( , ) ( , ) ( , )f x y g x y F u v G u v⇔
( , ) ( , ) ( , )* ( , )f x y g x y F u v G u v⇔
( , ) ( ) ( ) ( , ) ( ) ( )f x y f x f y F u v F u F v= ⇔ =
0 02 ( )0 0( , ) ( , )j ux vyf x x y x e F u vπ− +− − ⇔
0 02 ( )0 0( , ) ( , )j u x v ye f x y F u u v vπ + ⇔ − −
81
![Page 82: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/82.jpg)
Separability
1. Separability of the 2D Fourier transform – 2D Fourier Transforms can be implemented as a sequence of 1D Fourier
Transform operations performed independently along the two axis
( )
2 ( )
2 2 2 2
2
( , ) ( , )
( , ) ( , )
( , ) ,
j ux vy
j ux j vy j vy j ux
j vy
F u v f x y e dxdy
f x y e e dxdy e dy f x y e dx
F u y e dy F u v
π
π π π π
π
∞ ∞− +
−∞ −∞
∞ ∞ ∞ ∞− − − −
−∞ −∞ −∞ −∞
∞−
−∞
= =
= =
= =
∫ ∫
∫ ∫ ∫ ∫
∫
2D FT 1D FT along the rows
1D FT along the cols
82
![Page 83: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/83.jpg)
Separability
• Separable functions can be written as
2. The FT of a separable function is the product of the FTs of the two functions
( ) ( )
( ) ( )
2 ( )
2 2 2 2
( , ) ( , )
( ) ( )
j ux vy
j ux j vy j vy j ux
F u v f x y e dxdy
h x g y e e dxdy g y e dy h x e dx
H u G v
π
π π π π
∞ ∞− +
−∞ −∞
∞ ∞ ∞ ∞− − − −
−∞ −∞ −∞ −∞
= =
= =
=
∫ ∫
∫ ∫ ∫ ∫
( ) ( ) ( ) ( ) ( ) ( ), ,f x y h x g y F u v H u G v= ⇒ =
( ) ( ) ( ),f x y f x g y=
83
![Page 84: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/84.jpg)
Discrete Time Fourier Transform (DTFT)
Applies to Discrete domain signals and time series - 2D
84
![Page 85: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/85.jpg)
85
Fourier Transform: 2D Discrete Signals
■ Fourier Transform of a 2D discrete signal is defined as
where
2 ( )( , ) [ , ] j um vn
m nF u v f m n e π
∞ ∞− +
=−∞ =−∞
= ∑ ∑
1 1,2 2
u v−≤ <
1/ 2 1/ 22 ( )
1/ 2 1/ 2
[ , ] ( , ) j um vnf m n F u v e dudvπ +
− −
= ∫ ∫
■ Inverse Fourier Transform
![Page 86: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/86.jpg)
Properties
• Periodicity: 2D Fourier Transform of a discrete a-periodic signal is periodic – The period is 1 for the unitary frequency notations and 2π for normalized
frequency notations. – Proof (referring to the firsts case)
( )2 ( ) ( )( , ) [ , ] j u k m v l n
m nF u k v l f m n e π
∞ ∞− + + +
=−∞ =−∞
+ + = ∑ ∑
( )2 2 2[ , ] j um vn j km j ln
m nf m n e e eπ π π
∞ ∞− + − −
=−∞ =−∞
= ∑ ∑
2 ( )[ , ] j um vn
m nf m n e π
∞ ∞− +
=−∞ =−∞
= ∑ ∑
1 1
( , )F u v=
Arbitrary integers
86
![Page 87: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/87.jpg)
87
Fourier Transform: Properties
■ Linearity, shifting, modulation, convolution, multiplication, separability, energy conservation properties also exist for the 2D Fourier Transform of discrete signals.
![Page 88: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/88.jpg)
88
DTFT Properties
■ Linearity
■ Shifting
■ Modulation
■ Convolution
■ Multiplication
■ Separable functions
■ Energy conservation
[ , ] [ , ] ( , ) ( , )af m n bg m n aF u v bG u v+ ⇔ +
0 02 ( )0 0[ , ] ( , )j um vnf m m n n e F u vπ− +− − ⇔
[ , ] [ , ] ( , )* ( , )f m n g m n F u v G u v⇔
[ , ]* [ , ] ( , ) ( , )f m n g m n F u v G u v⇔
0 02 ( )0 0[ , ] ( , )j u m v ne f m n F u u v vπ + ⇔ − −
[ , ] [ ] [ ] ( , ) ( ) ( )f m n f m f n F u v F u F v= ⇔ =2 2[ , ] ( , )
m nf m n F u v dudv
∞ ∞∞ ∞
=−∞ =−∞ −∞ −∞
=∑ ∑ ∫ ∫
![Page 89: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/89.jpg)
89
Fourier Transform: Properties
■ Define Kronecker delta function
■ Fourier Transform of the Kronecker delta function
1, for 0 and 0[ , ]
0, otherwisem n
m nδ= =⎧ ⎫
= ⎨ ⎬⎩ ⎭
( ) ( )2 2 0 0( , ) [ , ] 1j um vn j u v
m nF u v m n e eπ πδ
∞ ∞− + − +
=−∞ =−∞
⎡ ⎤= = =⎣ ⎦∑ ∑
![Page 90: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/90.jpg)
90
Impulse Train
■ Define a comb function (impulse train) as follows
, [ , ] [ , ]M Nk l
comb m n m kM n lNδ∞ ∞
=−∞ =−∞
= − −∑ ∑
where M and N are integers
2[ ]comb n
n
1
![Page 91: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/91.jpg)
91
Impulse Train
combM ,N [m,n] δ[m− kM ,n− lN ]l=−∞
∞
∑k=−∞
∞
∑
[ ] 1, ,k l k l
k lm kM n lN u vMN M N
δ δ∞ ∞ ∞ ∞
=−∞ =−∞ =−∞ =−∞
⎛ ⎞− − ⇔ − −⎜ ⎟⎝ ⎠
∑ ∑ ∑ ∑
1 1,( , )
M N
comb u v, [ , ]M Ncomb m n
combM ,N (x, y) δ x − kM , y − lN( )l=−∞
∞
∑k=−∞
∞
∑
• Fourier Transform of an impulse train is also an impulse train:
![Page 92: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/92.jpg)
92
Impulse Train
2[ ]comb n
n u
1 12
12
1 ( )2comb u
12
![Page 93: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/93.jpg)
93
Impulse Train
( ) 1, ,k l k l
k lx kM y lN u vMN M N
δ δ∞ ∞ ∞ ∞
=−∞ =−∞ =−∞ =−∞
⎛ ⎞− − ⇔ − −⎜ ⎟⎝ ⎠
∑ ∑ ∑ ∑
1 1,( , )
M N
comb u v, ( , )M Ncomb x y
combM ,N (x, y) = δ x − kM , y − lN( )l=−∞
∞
∑k=−∞
∞
∑
• In the case of continuous signals:
![Page 94: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/94.jpg)
94
Impulse Train
2( )comb x
x u
1 12
12
1 ( )2comb u
12
2
![Page 95: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/95.jpg)
2D DTFT: constant
■ Fourier Transform of 1
To prove: Take the inverse Fourier Transform of the Dirac delta function and use the fact that the Fourier Transform has to be periodic with period 1.
f [k,l]=1,∀k,l
F[u,v]= 1× e− j2π uk+vl( )$%&
'()
l=−∞
∞
∑k=−∞
∞
∑ =
= δ(u− k,v − l)l=−∞
∞
∑k=−∞
∞
∑ periodic with period 1 along u and v
95
![Page 96: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/96.jpg)
Sampling theorem revisited
96
![Page 97: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/97.jpg)
97
Sampling
x
xM
( )f x
( )Mcomb x
u
( )F u
u
1( )* ( )M
F u comb u
u1M
1 ( )M
comb u
x
( ) ( )Mf x comb x
![Page 98: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/98.jpg)
98
Sampling
x
( )f x
u
( )F u
u
1( )* ( )M
F u comb u
x
( ) ( )Mf x comb x
WW−
M
W
1M1 2W
M>Nyquist theorem: No aliasing if
![Page 99: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/99.jpg)
99
Sampling
u
1( )* ( )M
F u comb u
x
( ) ( )Mf x comb x
M
W
1M
If there is no aliasing, the original signal can be recovered from its samples by low-pass filtering.
12M
![Page 100: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/100.jpg)
100
Sampling
x
( )f x
u
( )F u
u
1( )* ( )M
F u comb u
( ) ( )Mf x comb x
WW−
W
1MAliased
![Page 101: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/101.jpg)
101
Sampling
x
( )f x
u
( )F u
u[ ]( )* ( ) ( )Mf x h x comb x
WW−
1M
Anti-aliasing filter
uWW−
( )* ( )f x h x12M
![Page 102: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/102.jpg)
102
Sampling
u[ ]( )* ( ) ( )Mf x h x comb x
1M
u( ) ( )Mf x comb x
W
1M
■ Without anti-aliasing filter:
■ With anti-aliasing filter:
![Page 103: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/103.jpg)
103
Sampling in 2D (images)
x
y
u
v
uW
vW
x
y
( , )f x y ( , )F u v
M
N
, ( , )M Ncomb x y
u
v
1M
1N
1 1,( , )
M N
comb u v
![Page 104: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/104.jpg)
104
Sampling
u
v
uW
vW
,( , ) ( , )M Nf x y comb x y
1M
1N
1 2 uWM>No aliasing if and
1 2 vWN>
![Page 105: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/105.jpg)
105
Interpolation (low pass filtering)
u
v
1M
1N
1 1, for and v( , ) 2 2
0, otherwise
MN uH u v M N
⎧ ≤ ≤⎪= ⎨⎪⎩
12N
12M
Ideal reconstruction filter:
![Page 106: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/106.jpg)
106
Anti-Aliasing
a=imread(‘barbara.tif’);
![Page 107: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/107.jpg)
107
Anti-Aliasing
a=imread(‘barbara.tif’); b=imresize(a,0.25); c=imresize(b,4);
![Page 108: Fourier Transform · 2015. 3. 15. · • Continuous time a-periodic signal • Both time (space) and frequency are continuous variables – NON normalized frequency ω is used •](https://reader035.vdocuments.mx/reader035/viewer/2022071401/60ecbb00a758b40b305edbcd/html5/thumbnails/108.jpg)
108
Anti-Aliasing
a=imread(‘barbara.tif’); b=imresize(a,0.25); c=imresize(b,4); H=zeros(512,512); H(256-64:256+64, 256-64:256+64)=1; Da=fft2(a); Da=fftshift(Da); Dd=Da.*H; Dd=fftshift(Dd); d=real(ifft2(Dd));