Wavelet Transform - University of California, Berkeley
Transcript of Wavelet Transform - University of California, Berkeley
![Page 1: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/1.jpg)
Wavelet Transform
![Page 2: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/2.jpg)
4/14/2014 2
Wavelet Definition
“The wavelet transform is a tool that cuts up data, functions or operators into different frequency components, and then studies each component with a resolution matched to its scale”
Dr. Ingrid Daubechies, Lucent, Princeton U.
![Page 3: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/3.jpg)
4/14/2014 3
Fourier vs. Wavelet
► FFT, basis functions: sinusoids
► Wavelet transforms: small waves, called wavelet
► FFT can only offer frequency information
► Wavelet: frequency + temporal information
► Fourier analysis doesn’t work well on discontinuous, “bursty” data
music, video, power, earthquakes,…
![Page 4: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/4.jpg)
4/14/2014 4
Fourier vs. Wavelet
► Fourier
Loses time (location) coordinate completely
Analyses the whole signal
Short pieces lose “frequency” meaning
► Wavelets
Localized time-frequency analysis
Short signal pieces also have significance
Scale = Frequency band
![Page 5: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/5.jpg)
4/14/2014 5
Fourier transform
Fourier transform:
![Page 6: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/6.jpg)
4/14/2014 6
Wavelet Transform
► Scale and shift original waveform
► Compare to a wavelet
► Assign a coefficient of similarity
![Page 7: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/7.jpg)
4/14/2014 7
Scaling-- value of “stretch”
► Scaling a wavelet simply means stretching (or
compressing) it.
f(t) = sin(t)
scale factor1
![Page 8: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/8.jpg)
4/14/2014 8
Scaling-- value of “stretch”
► Scaling a wavelet simply means stretching (or
compressing) it.
f(t) = sin(2t)
scale factor 2
![Page 9: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/9.jpg)
4/14/2014 9
Scaling-- value of “stretch”
► Scaling a wavelet simply means stretching (or
compressing) it.
f(t) = sin(3t)
scale factor 3
![Page 10: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/10.jpg)
4/14/2014 10
More on scaling
► It lets you either narrow down the frequency band of interest, or determine the frequency content in a narrower time interval
► Scaling = frequency band
► Good for non-stationary data
► Low scalea Compressed wavelet Rapidly changing detailsHigh frequency
► High scale a Stretched wavelet Slowly changing, coarse features
Low frequency
![Page 11: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/11.jpg)
4/14/2014 11
Scale is (sort of) like frequency
Small scale
-Rapidly changing details,
-Like high frequency
Large scale
-Slowly changing
details
-Like low frequency
![Page 12: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/12.jpg)
4/14/2014 12
Scale is (sort of) like frequency
The scale factor works exactly the same with wavelets.
The smaller the scale factor, the more "compressed"
the wavelet.
![Page 13: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/13.jpg)
4/14/2014 13
Shifting
Shifting a wavelet simply means delaying (or hastening) its
onset. Mathematically, delaying a function f(t) by k is
represented by f(t-k)
![Page 14: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/14.jpg)
4/14/2014 14
Shifting
C = 0.0004
C = 0.0034
![Page 15: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/15.jpg)
4/14/2014 15
Five Easy Steps to a Continuous Wavelet
Transform
1. Take a wavelet and compare it to a section at the start of
the original signal.
2. Calculate a correlation coefficient c
![Page 16: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/16.jpg)
4/14/2014 16
Five Easy Steps to a Continuous Wavelet
Transform
3. Shift the wavelet to the right and repeat steps 1 and 2 until you've
covered the whole signal.
4. Scale (stretch) the wavelet and repeat steps 1 through 3.
5. Repeat steps 1 through 4 for all scales.
![Page 17: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/17.jpg)
4/14/2014 17
Coefficient Plots
![Page 18: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/18.jpg)
4/14/2014 18
Discrete Wavelet Transform
► “Subset” of scale and position based on power of two
rather than every “possible” set of scale and position in continuous wavelet transform
► Behaves like a filter bank: signal in, coefficients out
► Down-sampling necessary
(twice as much data as original signal)
![Page 19: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/19.jpg)
4/14/2014 19
Discrete Wavelet transform
signal
filters
Approximation
(a) Details
(d)
lowpass highpass
![Page 20: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/20.jpg)
4/14/2014 20
Results of wavelet transform — approximation and details
► Low frequency:
approximation (a)
► High frequency
details (d)
► “Decomposition”
can be performed
iteratively
![Page 21: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/21.jpg)
4/14/2014 21
Wavelet synthesis
•Re-creates signal from coefficients
•Up-sampling required
![Page 22: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/22.jpg)
4/14/2014 22
Multi-level Wavelet Analysis
Multi-level wavelet
decomposition tree Reassembling original signal
![Page 23: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/23.jpg)
4/14/2014 23
Subband Coding
![Page 24: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/24.jpg)
4/14/2014 24
2-D 4-band filter bank
Approximation
Vertical detail
Horizontal detail
Diagonal details
![Page 25: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/25.jpg)
4/14/2014 25
An Example of One-level Decomposition
![Page 26: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/26.jpg)
4/14/2014 26
An Example of Multi-level Decomposition
![Page 27: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/27.jpg)
4/14/2014 27
Wavelet Series Expansions
0 0
0
0
2
, ,
Wavelet series expansion of function ( ) ( )
relative to wavelet ( ) and scaling function ( )
( ) ( ) ( ) ( ) ( )
where ,
( ) : approximation and/or scaling coefficients
j j k j j k
k j j k
j
f x L
x x
f x c k x d k x
c k
d
( ) : detail and/or wavelet coefficientsj k
![Page 28: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/28.jpg)
4/14/2014 28
Wavelet Series Expansions
0 0 0, ,
, ,
( ) ( ), ( ) ( ) ( )
( ) ( ), ( ) ( ) ( )
j j k j k
j j k j k
c k f x x f x x dx
and
d k f x x f x x dx
![Page 29: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/29.jpg)
4/14/2014 29
Wavelet Transforms in Two Dimensions
( , ) ( ) ( )
( , ) ( ) ( )
( , ) ( ) ( )
( , ) ( ) ( )
H
V
D
x y x y
x y x y
x y x y
x y x y
/2
, ,
/2
, ,
( , ) 2 (2 ,2 )
( , ) 2 (2 ,2 )
{ , , }
j j j
j m n
i j i j j
j m n
x y x m y n
x y x m y n
i H V D
0
1 1
0 , ,
0 0
1( , , ) ( , ) ( , )
M N
j m n
x y
W j m n f x y x yMN
1 1
, ,
0 0
1( , , ) ( , ) ( , )
, ,
M Ni i
j m n
x y
W j m n f x y x yMN
i H V D
![Page 30: Wavelet Transform - University of California, Berkeley](https://reader031.fdocuments.in/reader031/viewer/2022012013/615865eb26695062e665a652/html5/thumbnails/30.jpg)
4/14/2014 30
Inverse Wavelet Transforms in Two Dimensions
0
0
0 , ,
, ,
, ,
1( , ) ( , , ) ( , )
1 ( , , ) ( , )
j m n
m n
i i
j m n
i H V D j j m n
f x y W j m n x yMN
W j m n x yMN