You are on page 1of 15

2D Continuous Wavelet Transform

Heejong Yoo(ECE)
April 26, 2001
Project Description

The idea of this project is from the software of Crit-


tech.com(Psilets 3.0) (www.crit-tech.com)

Project Description
Understand 1D, 2D DWT(FWT_PO.m, FWT2_PO.m in
WaveLab)
Understand 1D CWT(CWT.m in WaveLab)
Develop 2D CWT algorithm for image processing at fixed scale
CWT & DWT

CWT DWT
1. Scale At any scale Dyadic scales
2. Translation At any point Integer point
3. Wavelet Any wavelet that satisfies Orthogonal, biorthogonal,
minimum criteria
4. Computation Large Small
5. Detection Easily detects direction, Cannot detect minute
orientation object if not finely tuned
6. Application Pattern Recognition Compression
Feature extraction De-noising
Detection Transmission
Characterization
Continuous Wavelet Transform

In one dimension,
1 x b
f ( x) L2 ( R), cwt ( s , b )
s f ( x) (
s
)dx (time domain)

CWT ( s, w) s F ( w) ( sw) (frequency domain)

In two dimension,
1 x a y b
f ( x, y) L2 ( R), cwt ( s, a, b) f ( x, y) ( , )dxdy (time domain)
s s s
CWT ( s, w1, w2) s F ( w1, w2)( sw1, sw2) (frequency domain)

We want to calculate 2D CWT in frequency domain(just like WaveLab)


When scale s is fixed,
CWT ( w1, w2) F ( w1, w2)( w1, w2)
2D Mexican Hat wavelet

Time domain Frequency domain


1 1
( x2 y2 ) ( w12 w22 )
( x, y ) ( x y 2)e
2 2 2
(w1, w2) 2 (w1 w2 )e
2 2 2
2D Mexican Hat wavelet (Movie)
low frequency high frequency

<Time Domain Wavelet> <Fourier Domain Wavelet>


(w1 , w2 ) F{( x, y)}

1
( x2 y2 )
( w1 , w2 ) F { ( x, y )} [( x y 2)e 2 2 2
]e jw1 x e jw2 y dxdy
1 1 1 1
x2 y2 y2 x2
(x e 2 2
e jw1 x
)(e 2
e jw2 y
)dxdy ( y e 2 2
e jw2 y
)(e 2
e jw1 x )dydx
1 1
x2 y2
2 (e 2
e jw1 x
)(e 2
e jw2 y )dxdy
1 1 1 1
x2 y2 y2 x2
(x e
2 2
e jw1 x
)dx (e 2
e jw2 y
)dy ( y e 2 2
e jw2 y
)dy (e 2
e jw1 x )dx
1 1
x2 y2
2 (e 2
e jw1 x
)dx (e 2
e jw2 y )dy
1 1 1 1
w12 w2 2 w2 2 w12
( 2 (1 w )e 2
1
2
)( 2 e 2
) ( 2 (1 w )e 2
2
2
)( 2 e 2
)
1 1
w12 w2 2
2( 2 e 2
)( 2 e 2
)
1
( w12 w22 )
2 ( w w )e
2
1
2
2
2
Output Image Display
GUI based 2D CWT 4 different Input Image

In general,
low scale means high frequency,
high scale means low frequency

In WaveLab,
low scale means low frequency,
high scale means high frequency

Wavelet Select
This program follows the WaveLab

Scale(1~100)

Fourier Domain Wavelet


Crit-tech Psilets 3.0 Output (1)
Scale = 38
Scale =2
Scale =1
Crit-tech Psilets 3.0 Output (2)
Scale =48
Scale =3

You might also like