2-D Orthogonal Transforms

2-D Discrete Fourier Transform (DFT)

Recapitulation

TSC 2002 Digital Image Processing

Lecture #4

Dr K.C.Raveendranathan SM IEEE
FIE & C.Eng, FIETE, LM ISTE, LM CSI and LM IRSS

LBS Institute of Technology for Women Poojappura

Department of Electronics and Communication Engineering

May 7, 2014

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Outline of the Lecture

1 2-D Orthogonal Transforms

2 2-D Discrete Fourier Transform (DFT)

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Outline of the Lecture

1 2-D Orthogonal Transforms

2 2-D Discrete Fourier Transform (DFT)

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Introduction
In general, transforms such as Fourier, Laplace, and
Z-transform obtain the frequency content of signals. In other
words, the transformation operations switches the temporal
representation of signals to that of frequency domain
representation. We begin our discussion with Fourier Transform
pairs:
Z Z∞
F{f (x, y )} = F (u, v ) = f (x, y )exp[−j2π(ux + vy )]dxdy (1)
−∞
Z Z∞
−1
F {F (u, v )} = F (u, v )exp[j2π(ux + vy )]dudv (2)
−∞

What are the conditions for these transforms?

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM
TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

2-D Fourier Transforms–contd

Z Z∞
F{f (x, y )} = F (u, v ) = f (x, y )exp[−j2π(ux + vy )]dxdy
−∞
Z Z∞
F −1 {F (u, v )} = F (u, v )exp[j2π(ux + vy )]dudv
−∞

What the conditions for these transforms to exist?

f(x,y) is continuous and integrable.
F(u,v) is integrable.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

2-D Fourier Transforms–contd

Z Z∞
F{f (x, y )} = F (u, v ) = f (x, y )exp[−j2π(ux + vy )]dxdy
−∞
Z Z∞
F −1 {F (u, v )} = F (u, v )exp[j2π(ux + vy )]dudv
−∞

What the conditions for these transforms to exist?

f(x,y) is continuous and integrable.
F(u,v) is integrable.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

2-D Fourier Transforms–contd

Z Z∞
F{f (x, y )} = F (u, v ) = f (x, y )exp[−j2π(ux + vy )]dxdy
−∞
Z Z∞
F −1 {F (u, v )} = F (u, v )exp[j2π(ux + vy )]dudv
−∞

What the conditions for these transforms to exist?

f(x,y) is continuous and integrable.
F(u,v) is integrable.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Fourier Spectrum, Phase, and Power Spectrum

The Fourier Spectrum, phase, and power spectrum are given

by:
h i
|F (u, v )| = R 2 (u, v ) + I 2 (u, v ) (3)
 
I(u, v )
φ(u, v ) = tan−1 (4)
R(u, v )
P(u, v ) = |F (u, v )|2 = R 2 (u, v ) + I 2 (u, v ) (5)

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Discrete Fourier Transform (DFT)

The function f (x, y ) is discretized into a sequence:

f (x0 , y0 ), f (x0 + ∆x, y0 + ∆y ), . . . ,

f (x0 + [N − 1]∆x, y0 + [N − 1]∆y )

where (x, y ) ∈ Z × Z . viz. x and y take discrete values. In

otherwords, the sequence, {f (0, 0), f (1, 1), . . . , f (N − 1, N − 1)}
denotes any N uniformly spaced samples from a corresponding
continuous function.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Discrete Fourier Transform (DFT)-contd.

The DFT pair is given by:

M−1 N−1
1 XX
F (u, v ) = f (x, y ) exp[−j2π(ux/M + vy /N)](6)
MN
x=0 y =0
M−1
X N−1
X
f (x, y ) = F (u, v ) exp[j2π(ux/M + vy /N)] (7)
u=0 v =0

Question:-(a) How to compute the Discrete Fourier Transform

(DFT) of a 2-D matrix in MATLAB?
(b) Which algorithm is implemented for (a) above?

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Discrete Fourier Transform (DFT)-contd.

The discrete function f (x, y ) represents samples of the function

f (x0 + x∆x, y0 + y ∆y ), for x = 0, 1, 2, . . . , M − 1 and
y = 0, 1, 2, . . . , N − 1 . The sampling increments in the spatial
and frequency domains are related by:

1
∆u = , and (8)
M∆x
1
∆v = (9)
N∆y

Question:- What will be change to Equations (6) and (7) when

images are sampled in a square array?

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Reading Assignments & Home work

Chapter 2 of Anil K. Jain, Fundamentals of DIP.

Chapter 3 of Gonzalez & Woods, DIP.
Refresh the basics of LTI systems, convolution, and
properties of common transforms.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Reading Assignments & Home work

Chapter 2 of Anil K. Jain, Fundamentals of DIP.

Chapter 3 of Gonzalez & Woods, DIP.
Refresh the basics of LTI systems, convolution, and
properties of common transforms.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Reading Assignments & Home work

Chapter 2 of Anil K. Jain, Fundamentals of DIP.

Chapter 3 of Gonzalez & Woods, DIP.
Refresh the basics of LTI systems, convolution, and
properties of common transforms.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Recap

We considered the following in this lecture:

2-D Transforms–Fourier Transforms, Conditions for
existence..
2-D Discrete Fourier Transform (DFT)
Reading Assignments.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Recap

We considered the following in this lecture:

2-D Transforms–Fourier Transforms, Conditions for
existence..
2-D Discrete Fourier Transform (DFT)
Reading Assignments.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing
 2-D Orthogonal Transforms
2-D Discrete Fourier Transform (DFT)
Recapitulation

Recap

We considered the following in this lecture:

2-D Transforms–Fourier Transforms, Conditions for
existence..
2-D Discrete Fourier Transform (DFT)
Reading Assignments.

Dr K.C.Raveendranathan SM IEEE FIE & C.Eng, FIETE, LM ISTE, LM

TSC
CSI2002
and Digital
LM IRSS Image Processing