Presents

Robi Polikar Dept. of Electrical & Computer Engineering Rowan University

The Story of Wavelets

Technical Overview
ButWe cannot do that with Fourier Transform. Time - frequency representation and the STFT Continuous wavelet transform Multiresolution analysis and discrete wavelet transform (DWT)

Application Overview
Conventional Applications: Data compression, denoising, solution of PDEs, biomedical signal analysis. Unconventional applications YesWe can do that with wavelets too

Historical Overview
1807 ~ 1940s: The reign of the Fourier Transform 1940s ~ 1970s: STFT and Subband Coding 1980s & 1990s: The Wavelet Transform and MRA

What is a Transform and Why Do we Need One ?

Transform: A mathematical operation that takes a function or sequence and maps it into another one Transforms are good things because The transform of a function may give additional /hidden information about the original function, which may not be available /obvious otherwise The transform of an equation may be easier to solve than the original equation (recall your fond memories of Laplace transforms in DFQs) The transform of a function/sequence may require less storage, hence provide data compression / reduction An operation may be easier to apply on the transformed function, rather than the original function (recall other fond memories on convolution).

December, 21, 1807

An arbitrary function, continuous or with discontinuities, defined in a finite interval by an arbitrarily capricious graph can always be expressed as a sum of sinusoids J.B.J. Fourier Jean B. Joseph Fourier (1768-1830)

Complex Function ! weight i y Simple Function i

i

Complex function representation through simple building blocks

Basis functions

Using only a few blocks Compressed representation Using sinusoids as building blocks Fourier transform
Frequency domain representation of the function

([) ! f (t)e

 j[t

dt

1 f (t) ! ([)e j[td[ 2T

How Does FT Work Anyway?

Recall that FT uses complex exponentials (sinusoids) as building blocks. j[ t
e ! cos [ t  j sin [ t

For each frequency of complex exponential, the sinusoid at that frequency is compared to the signal. If the signal consists of that frequency, the correlation is high large FT coefficients.
([ )

f (t ) e

 j[t

dt

f (t )

1 2T

([ )e j[t d[

If the signal does not have any spectral component at a frequency, the correlation at that frequency is low / zero, small / zero FT coefficient.

FT At Work

x1 (t ) ! cos(2T 5 t )

x2 (t ) ! cos(2T 25 t )

x3 (t ) ! cos(2T 50 t )

FT At Work

x1 (t )

X 1 ([ )

x2 (t )

X 2 ([ )

x3 (t )

X 3 ([ )

FT At Work

x4 (t ) ! cos(2T 5 t )  cos(2T 25 t )  cos(2T 50 t )

x 4 (t )

X 4 ([ )

FT At Work
Complex exponentials (sinusoids) as basis functions:
g

F ([ ) !

f (t ) e
g

An ultrasonic A-scan using 1.5 MHz transducer, sampled at 10 MHz

1 f (t ) ! 2T

F ([ ) e
g

j[t

dt
j[t

dt

Stationary and Non-stationary Signals

FT identifies all spectral components present in the signal, however it does not provide any information regarding the temporal (time) localization of these components. Why? Stationary signals consist of spectral components that do not change in time all spectral components exist at all times no need to know any time information FT works well for stationary signals However, non-stationary signals consists of time varying spectral components How do we find out which spectral component appears when? FT only provides what spectral components exist , not where in time they are located. Need some other ways to determine time localization of spectral components

Stationary and Non-stationary Signals

Stationary signals spectral characteristics do not change with time
x4 (t ) ! cos(2T 5 t )  cos(2T 25 t )  cos(2T 50 t )

Non-stationary signals have time varying spectra

x5 (t ) ! [ x1 x2 x3 ]
Concatenation

Stationary vs. Non-Stationary

X4( )

Perfect knowledge of what frequencies exist, but no information about where X5( ) these frequencies are located in time

Shortcomings of the FT
Sinusoids and exponentials
Stretch into infinity in time, Instantaneous in frequency, no time localization perfect spectral localization

Global analysis does not allow analysis of non-stationary signals

Need a local analysis scheme for a time-frequency representation (TFR) of nonstationary signals
Windowed F.T. or Short Time F.T. (STFT) : Segmenting the signal into narrow time intervals, narrow enough to be considered stationary, and then take the Fourier transform of each segment, Gabor 1946. Followed by other TFRs, which differed from each other by the selection of the windowing function

Short Time Fourier Transform (STFT)

1. 2. 3. 4. 5. 6. Choose a window function of finite length Place the window on top of the signal at t=0 Truncate the signal using this window Compute the FT of the truncated signal, save. Incrementally slide the window to the right Go to step 3, until window reaches the end of the signal For each time location where the window is centered, we obtain a different FT
Hence, each FT provides the spectral information of a separate time-slice of the signal, providing simultaneous time and frequency information

STFT

Time parameter

Frequency parameter

Signal to be analyzed

FT Kernel (basis function)

[ d STFTx (t , [ ) !

)A ?x(t ) W (t  t d e
t

j [t

dt

STFT of signal x(t): Computed for each window centered at t=t

Windowing function

Windowing function centered at t=t

STFT
t=t=-8 t=-2 t=-

t=4

t=8

STFT at Work

STFT At Work

STFT At Work

STFT

STFT provides the time information by computing a different FTs for consecutive time intervals, and then putting them together
Time-Frequency Representation (TFR) Maps 1-D time domain signals to 2-D time-frequency signals

Consecutive time intervals of the signal are obtained by truncating the signal using a sliding windowing function How to choose the windowing function?
What shape? Rectangular, Gaussian, Elliptic? How wide?
 Wider window require less time steps low time resolution  Also, window should be narrow enough to make sure that the portion of the signal falling within the window is stationary  Can we choose an arbitrarily narrow window?

Selection of STFT Window

[ STFTx (t d ) ! ?x(t ) W (t  t d e  j[t dt ,[ )A t
Two extreme cases: W(t) infinitely long: W (t ) ! 1 STFT turns into FT, providing excellent frequency information (good frequency resolution), but no time information W(t) infinitely short: W (t ) ! H (t )
[ STFTx (t d ) ! ?x(t ) H (t  t d e  j[t dt ! x(t d e  j[t d ,[ )A ) t

STFT then gives the time signal back, with a phase factor. Excellent time information (good time resolution), but no frequency information Wide analysis window poor time resolution, good frequency resolution Narrow analysis window good time resolution, poor frequency resolution Once the window is chosen, the resolution is set for both time and frequency.

Heisenberg Principle

1 (t (f u 4T

Time resolution: How well two spikes in time can be separated from each other in the transform domain

Frequency resolution: How well two spectral components can be separated from each other in the transform domain

Both time and frequency resolutions cannot be arbitrarily high!!!

We cannot precisely know at what time instance a frequency component is located. We can only know what interval of frequencies are present in which time intervals

The Wavelet Transform

Overcomes the preset resolution problem of the STFT by using a variable length window Analysis windows of different lengths are used for different frequencies:

Analysis of high frequencies Use narrower windows for better time resolution Analysis of low frequencies Use wider windows for better frequency resolution
This works well, if the signal to be analyzed mainly consists of slowly varying characteristics with occasional short high frequency bursts. Heisenberg principle still holds!!! The function used to window the signal is called the wavelet

The Wavelet Transform

A normalization Translation parameter, Scale parameter, Signal to be measure of time measure of frequency constant analyzed

1 ] ] CW x (X , s ) ! =x (X , s) !

s t

] xt

 t X

dt s

Continuous wavelet transform of the signal x(t) using the analysis wavelet ](.)

The mother wavelet. All kernels are obtained by translating (shifting) and/or scaling the mother wavelet

Scale = 1/frequency

WTx (X , s ) ! =x (X , s ) !

t X x t  dt s s t

WT at Work

Low frequency (large High frequency (small scale) scale)

WT at Work

WT at Work

WT at Work

Matlab Demos on CWT

Discrete Wavelet Transform

CWT computed by computers is really not CWT, it is a discretized version of the CWT. The resolution of the time-frequency grid can be controlled (within Heisenbergs inequality), can be controlled by time and scale step sizes. Often this results in a very redundant representation How to discretize the continuous time-frequency plane, so that the representation is non-redundant?
Sample the time-frequency plane on a dyadic (octave) grid
1 t X ] ] CW x (X , s) ! =] (X , s ) ! xt  x s s t
dt

] kn (t ) ! 2  k ] 2  k  n

k, n Z

Discrete Wavelet Transform

Dyadic sampling of the time frequency plane results in a very efficient algorithm for computing DWT:
Subband coding using multiresolution analysis Dyadic sampling and multiresolution is achieved through a series of filtering and up/down sampling operations

x[n]

H
N

y[n]

y[n ] ! x[n ] * h[n ] ! h[n ] * x[n ] ! x[k ] h[n  k ]

k !1 N

! h[k ] x[n  k ]
k !1

Discrete Wavelet Transform Implementation

x[n] ~ G ~ H
~

yhigh [k ] ! x[n] g[ n  2 k ]

n

yhigh [k ] g[n  2k ]
k

x[n]

2 2

G H

~ G ~ H

2 2

2 2

G H

ylo [k ] ! x[n] h[ n  2 k ]

yhigh [k ] g[n  2k ]
k

Decomposition
G H

Reconstruction
2 Down-sampling 2 Up-sampling

Half band high pass filter Half band low pass filter

2-level DWT decomposition. The decomposition can be continues as long as there are enough samples for down-sampling. down-

DWT - Demystified
Length: 512 B: 0 ~ T

|H(jw)|

g[n]
Length: 256 B: T/2 ~ T Hz

h[n]
2

2
d1: Level 1 DWT Coeff.

a1

Length: 256 B: 0 ~ T/2 Hz

-T/2

T/2

|G(jw)|

g[n]
2

h[n]
2
Length: 128 B: 0 ~ T /4 Hz

Length: 128 B: T/4 ~ T/2 Hz

a2
d2: Level 2 DWT Coeff.

-T

-T/2

T/2

g[n]
2
d3: Level 3 DWT Coeff.

h[n]
2
Length: 64 B: 0 ~ T/8 Hz

Length: 64 B: T/8 ~ T/4 Hz

a3.Level 3 approximation
Coefficients

Implementation of DWT on MATLAB

Choose wavelet and number of levels Hit Analyze button

Load signal s=a5+d5++d1 Approx. coef. at level 5

Level 1 coeff. Highest freq.

(Wavedemo_signal1)

Applications of Wavelets

Compression De-noising Feature Extraction Discontinuity Detection Distribution Estimation Data analysis
Biological data NDE data Financial data

Compression
DWT is commonly used for compression, since most DWT are very small, can be zeroed-out!

Compression

Compression

ECG- Compression

Denoising Implementation in Matlab

First, analyze the signal with appropriate wavelets

Hit Denoise

(Noisy Doppler)

Denoising Using Matlab

Choose thresholding method Choose noise type

Choose thrsholds

Hit Denoise

Denosing Using Matlab

Discontinuity Detection

(microdisc.mat)

Discontinuity Detection with CWT

(microdisc.mat)

Application Overview
Data Compression Wavelet Shrinkage Denoising Source and Channel Coding Biomedical Engineering EEG, ECG, EMG, etc analysis MRI Nondestructive Evaluation Ultrasonic data analysis for nuclear power plant pipe inspections Eddy current analysis for gas pipeline inspections Numerical Solution of PDEs Study of Distant Universes Galaxies form hierarchical structures at different scales

Application Overview

Wavelet Networks Real time learning of unknown functions Learning from sparse data Turbulence Analysis Analysis of turbulent flow of low viscosity fluids flowing at high speeds Topographic Data Analysis Analysis of geo-topographic data for reconnaissance / object identification Fractals Daubechies wavelets: Perfect fit for analyzing fractals Financial Analysis Time series analysis for stock market predictions

History Repeats Itself

1807, J.B. Fourier:

All periodic functions can be expressed as a weighted sum of trigonometric function Denied publication by Lagrange, Legendre and Laplace 1822: Fouriers work is finally published 143 years 1965, Cooley & Tukey: Fast Fourier Transform

History Repeats Itself: Morlets Story

1946, Gabor: STFT analysis: high frequency components using a narrow window, or low frequency components using a wide window, but not both Late 1970s, Morlets (geophysical engineer) problem: Time - frequency analysis of signals with high frequency components for short time spans and low frequency components with long time spans STFT can do one or the other, but not both Solution: Use different windowing functions for sections of the signal with different frequency content Windows to be generated from dilation / compression of prototype small, oscillatory signals wavelets Criticism for lack of mathematical rigor !!! Early 1980s, Grossman (theoretical physicist): Formalize the transform and devise the inverse transformation First wavelet transform ! Rediscovery of Alberto Calderons 1964 work on harmonic analysis

1980s

1984, Yeves Meyer :

Similarity between Morlets and Colderons work, 1984 Redundancy in Morlets choice of basis functions 1985, Orthogonal wavelet basis functions with better time and frequency localization

Rediscovery of J.O. Strombergs 1980 work the same basis functions (also a harmonic analyst) Yet re-rediscovery of Alfred Haars work on orthogonal basis functions, 1909 (!).
Simplest known orthonormal wavelets

Transition to the Discrete Signal Analysis

Ingrid Daubechies: Discretization of time and scale parameters of the wavelet transform Wavelet frames, 1986 Orthonormal bases of compactly supported wavelets (Daubechies wavelets), 1988 Liberty in the choice of basis functions at the expense of redundancy Stephane Mallat: Multiresolution analysis w/ Meyer, 1986 Ph.D. dissertation, 1988 Discrete wavelet transform Cascade algorithm for computing DWT

However

Decomposition of a discrete into dyadic frequencies (MRA) , known to EEs under the name of Quadrature Mirror Filters, Croisier, Esteban and Galand, 1976 (!)

Transition to the Discrete Signal Analysis

Martin Vetterli & Jelena Kovacevic Wavelets and filter banks, 1986 Perfect reconstruction of signals using FIR filter banks, 1988 Subband coding Multidimensional filter banks, 1992

1990s

Equivalence of QMF and MRA, Albert Cohen, 1990 Compactly supported biorthogonal wavelets, Cohen, Daubechies, J. Feauveau, 1993 Wavelet packets, Coifman, Meyer, and Wickerhauser, 1996 Zero Tree Coding, Schapiro 1993 ~ 1999 Search for new wavelets with better time and frequency localization properties. Super-wavelets Matching Pursuit, Mallat, 1993 ~ 1999

New & Noteworthy

Zero crossing representation signal classification computer vision data compression denoising Super wavelet Linear combination of known basic wavelets Zero Tree Coding, Schapiro Matching Pursuit , Mallat Using a library of basis functions for decomposition New MPEG standard