Wavelettour

A wavelet tutorial from S.
Mallat's book
A WAVELET TOUR
OF SIGNAL PROCESSING
BY STÉPHANE MALLAT
Academic Press, 1998
A SHORT PRESENTATION BY F.
CHAPLAIS
for those who hate preambles
Hardware and Software Requirements
This site has been tested with Netscape Navigator 2 and 3. Version 2 fails to render
characters from the Symbol font, but this does does seriously affect the readability of
the presentation. A 256 gray level display is required to view the illustrations. Some
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Wavetour_presentation_US.html (1 of 4) [11/15/2003 9:42:50 PM]

A wavelet tutorial from S. Mallat's book
graphs use color plots for better reading. Most images are interlaced GIF with
transparency color set to white.
Warning
This presentation is inspired from S.G. Mallat's book and does not pretend to reflect it
exactly. It is concerned with the following topics:
● Fourier analysis (chapter 2)

● time-frequency analysis (chapter 4, except for the quadratic
energy distributions))
● frames (chapter 5)
● singularity analysis and reconstruction (chapter 6 except for the
multifractals)
● wavelet bases and filter banks (chapter 7)
The following topics from the book are not covered here:
● chapter 3 about discrete signals (except for the FFT and

convolutions algorithms, which are briefly described)
● chapter 8 on wavelet packets and local cosine bases
● chapter 9 on approximation
● chapter 10 about estimation (which is being revised by S.G.
Mallat for the French edition and the second US edition)
● chapter 11 on compression and coding (hope to do it someday)
Proposed Tours
Four tours are proposed, corresponding to four different topics.

These tours are linked to each other sequentially. Many links allow
navigation from one to topic to another for a nonlinear browsing.
● Frequency Analysis
● From dyadic wavelets to filters
● From filters to dyadic wavelets
● Regularity Analysis
● Frames
Index
For direct access, here is a list of links that point to the main topics:
Fourier Transform
Instantaneous Frequency of an Analytic Signal
Time-Frequency Localization
Windowed Fourier Transform and Wavelet Transform
Frames and Riesz Bases
Windowed Fourier Frames and Wavelet Frames
Multiresolution Approximations
Wavelet Bases
Filter Banks
Regularity Analysis of a Signal
Detection of Singularities
Reconstruction from Dyadic Maxima
Edge Detection and Image Reconstruction
Algorithms:
Fast Fourier Transforms and Convolutions

Fast Windowed Fourier Transform
Fast Wavelet Transform
Fast Dyadic Wavelet Transform
Decomposition and Reconstruction over Orthonormal
Wavelet Bases
Numerical Computations
All numerical figures in the book have been computed using Wavelab, a freeware
Matlab Toolbox, available at
http://www-stat.stanford.edu/~wavelab/
Uvi Wave is another freeware Matlab Toolbox.

Updates:
June 4, 1999: Full Strang and Fix conditions added, with proof!
May 18, 2001: added mathematical transition from filter banks to multiresolution
analysis
Last update: February 2, 1999
Feedback is welcome.
Control and
Systems
Department

Frequency and Period
Frequency Analysis
As shown in this tour, the notion of instantaneous frequency should
be handled with caution. This is why we recall the primary
definition of a frequency.
Note: this presentation is proper to the site and does not come from the book.
Period and Frequency

What is a frequency?
A frequency is the inverse of a period.
A one dimensional signal is T>0 periodic if it is unchanged by a

translation of T. Hence its support has an infinite length.
Nonetheless, the signal is entirely determined by its values over an
interval of length T.
In the representation over an interval of length T, the

regularity of the signal implies the equality of the values of the
functions and of its derivatives at the left and right ends of the
interval.
A musical example
A purely synthetic music note can be represented by a sinusoidal
wave. An instrumental note that is held is a more complex signal.
A music or a speech recording is even more complex; in particular,
the frequencies may vary with time.
As a first approximation, a music piece may be modelized as a

sequence of signals defined over intervals whose length is
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Intro/Periodicity_US.html (1 of 2) [11/15/2003 9:42:52 PM]

Frequency and Period
determined by the tempo. Every elementary signal can be

periodized.
Fourier Analysis
Classically, the analysis of a signal as a Fourier series or a Fourier
integral provides a representation of its frequency contents.
Fourier series
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Intro/Periodicity_US.html (2 of 2) [11/15/2003 9:42:52 PM]

Fourier Series
Fourier Series
An N-periodic signal can be represented as a
series of harmonically related sinusoids.
Fourier has shown that any periodic signal can be decomposed into
harmonically related sinusoids:
This analysis seems to fits the needs of music coding by using a

Fourier series for every tempo unit. It has several drawbacks:
● there may be points where the Fourier series does not

converge
● a periodic continuous signal with a period different from 2π is
represented as a discontinuous signal over . This causes
a bad convergence of its Fourier series.
● More generally, it cannot efficiently represent signals which
are not synchronized with a tempo, such as speech recordings
or images.
A solution consists in using a transform where the frequency is

allowed to have continuous variations. The Fourier series becomes
the Fourier transform. The concept of period disappears in the
process, and the frequency contents is actually a sinusoid contents.
Fourier transform
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Fourier_Series.html [11/15/2003 9:42:54 PM]

Fourier transform
Fourier Transform
The Fourier transform analyses the "frequency
contents" of a signal.
Its many properties make it suitable for studying
linear time invariant operators, such as
differentiation.
It is a global representation of a signal.
Fourier transform
The Fourier transform of f in L2 is
The inverse Fourier transform represents f as a sum of sinusoids
Properties
The Fourier transform has many algebraic properties. Note that
sinusoidal waves are eigenvectors of the differentiation operator.
This makes it possible for the Fourier transform to give indications
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Fourier_Transform.html (1 of 2) [11/15/2003 9:42:57 PM]

Fourier transform
on the regularity of a signal.
Implementation
To reduce the number of operations, the Fast Fourier Transform
separates odd and even frequencies when computing a discrete
Fourier transform.
A global representation
The Fourier transform is a global representation of the signal. It
cannot analyze it local frequency contents or its local regularity.
The convergence condition on the Fourier transform only gives the
worst order of regularity. It ignores local regular behaviours.
There exist, however, a definition of the instantaneous frequency of

an analytic signal.
It is useless in practice because it fails to detect the summation of

two signals. It is nonetheless a convenient means of defining
synthetic signals for numerical experimentations.
Trying to discriminate each of the stacked frequencies leads to a

frequency analysis that is localized in frequency as well as in time.
This requires some understanding of the time-frequency
localization of signal.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Fourier_Transform.html (2 of 2) [11/15/2003 9:42:57 PM]

Algebraic properties of the Fourier transform
Algebraic Properties
of the Fourier Transform
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Agebraic_Properties.html [11/15/2003 9:43:00 PM]

Regularity analysis by means of the Fourier transform
Regularity Analysis
by Means of the Fourier
Transform
A function f is bounded and p times continuously differentiable
with bounded derivatives if
This property is extended to Lipschitz regularity:
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Regularity_with_Fourier.html [11/15/2003 9:43:13 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Regularite/LipschitzDefUS.GIF
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Regularite/LipschitzDefUS.GIF [11/15/2003 9:43:15 PM]

Fast Fourier Transform

The discrete Fourier transform of a discrete signal with N samples
is
Using this formula to compute the Fourier transform requires N2

complex additions and multiplications.
A simple computation shows that the even frequency coefficients

are the coefficients of the Fourier transform of the N/2 periodic
signal
fp[n] = f[n] + f[n+N/2]
and that the odd frequency coefficients are the coefficients of the
Fourier transform of
fi[n] = (f[n]-f[n+N/2]) e-2iπn/N
One verifies by induction that the number of operations required by

this method to compute the Fourier transform is of the order of KN
log2(N), where K is a constant which does not depend on N.
This is the basic principle of the Fast Fourier Transform. Variants

exist that reduce K.
Convolutions and circular convolutions

The circular convolution of two N periodic signals is defined by
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/FFTUS.html (1 of 2) [11/15/2003 9:43:21 PM]

The discrete Fourier transform of a circular convolution is the

product of the two discrete Fourier transforms.
For two signals f and h with a length M>=32, computing their

convolution with an FFT is faster than using the straightforward
formula. For that purpose, two M periodic signals are defined:
and one can verify that their circular convolution is equal to the
convolution
for 0<=n<2M. The circular convolution is itself computed by

performing the FFT of the two signals, computing their product
and then its inverse FFT.
Fourier Transform
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/FFTUS.html (2 of 2) [11/15/2003 9:43:21 PM]

Instantanous Frequency of an Analytic Signal
Instantaneous Frequency
of an Analytic Signal
A signal in L2 is analytic if its Fourier transform is zero for
negative frequencies. The analytic part fa of a real signal f is given
by its Fourier transform:
and fa can properly be decomposed into a module and a complex

phase:
The instantanneous frequency is the nonnegative derivative of the

complex phase:
For a sinusoidal wave, this definition coincides with the usual

frequency. Unfortunately, the instantaneous frequency of the sum
of two ordinary sinusoidal waves is the average of their
frequencies, which does not coincide with the result of a Fourier
analysis.
Hence the caracterization of the instaneous frequency of a signal in

a sense that is consistent with the Fourier analysis in simple cases
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Intro/Instantanous_Frequency.html (1 of 2) [11/15/2003 9:43:29 PM]

Instantanous Frequency of an Analytic Signal
requires other mathematical tools.
However, analytic signals are very useful in the synthesis of signals

with a given time varying frequency contents.
Fourier transform
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Intro/Instantanous_Frequency.html (2 of 2) [11/15/2003 9:43:29 PM]

Time-Frequency Atoms
There is no finite energy function which is
compactly supported both in the time and
frequency domains.
The time-frequency localization is measured in the
mean squares sense and is represented as a
Heisenberg box.
The Fourier transform can be viewed as a representation of a

function as a sum of sinusoidal waves. These sinusoids are very
well localized in the frequency, but not in time, since their support
has an infinite length. This is a consequence of periodicity.
To represent the frequency behavior of a signal locally in time, the

signal should be analyzed by functions which are localized both in
time and frequency, for instance, signals that are compactly
supported in the time and Fourier domains.
This time-frequency localization is limited by the following two

results:
The uncertainty theorem of Heisenberg.

If f is in L2, then its time root deviation and its Fourier root
deviation are defined. Then
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/atomes/Time-frequency_atoms.html (1 of 3) [11/15/2003 9:43:34 PM]

A balance has to be reached between the time and frequency

resolution. In the limit case of a sinusoid, is zero and is
infinite.
The previous inequality is an equality if and only if f is a Gabor

chirp .
Compact supports
If f is non zero with a compact support, then its Fourier transform
cannot be zero on a whole interval. Similarly, if its Fourier
transform is compactly supported, then it cannot be zero on a time
interval.
Hence, even if the Heisenberg constraints are verified, it is

impossible to have an function in L2 which is compactly supported
both in the time and Fourier domains.
In particular, this means that there is no instantaneous frequency

analysis for finite energy signals.
Time-frequency localization is thus

achievable only in the mean squares
sense.
This localization is represented as a Heisenberg box.

For a family of vectors to be a basis of L2, it is reasonable to

expect that their Heisenberg boxes pave the time frequency plane.
Two time frequency localization strategies are presented in

parallel; the first one leads to the windowed Fourier transform,
while the other one leads to the wavelet transform.
Windowed Fourier transforms and wavelet transforms

Fast Windowed Fourier

Transform
The discrete windowed Fourier transform of an N periodic signal is
For a fixed m, the formula is that of a discrete Fourier transform.

Hence the computation is performed with N fast Fourier
transforms, which require O(N2 log2(N)) operations.
Windowed Fourier Transform
http://cas.ensmp.fr/~chaplais/Wavetour_presentation...sformees/Fourier_Fenetre/Fast_Windowed_Fourier.html [11/15/2003 9:43:35 PM]


The windowed Fourier transform replaces the
Fourier transform's sinusoidal wave by the
product of a sinusoid and a window which is
localized in time.
It takes two arguments: time and frequency.
Outline
The windowed Fourier transform is defined by
It uses an atom which is the product of a sinusoidal wave with a

finite energy symmetric window g. The windowed Fourier
transform family of atoms is obtained by time translations and
frequency modulations of the original window:
This atom has a frequency center ξ and is symmetric with respect

to u.
The time and frequency spreads of these functions are constant.

The family is generated by time and frequency translations of one
atom. Here are examples of Heisenberg boxes of windowed Fourier
http://cas.ensmp.fr/~chaplais/Wavetour_presenta...nsformees/Fourier_Fenetre/Windowed_Fourier.html (1 of 3) [11/15/2003 9:43:40 PM]

atoms:
Properties
The windowed Fourier transform has a constant time frequency
resolution. This resolution can be can be changed by rescaling the
window g. It is a complete, stable, redundant representation of the
signal. Hence it is invertible. The redundancy implies the existence
of a reproducing kernel.
Spectrogram
The square modulus of the windowed Fourier transform is the
spectrogram of a signal:
Choice of Window
The properties of the windowed Fourier transform are determined

by the window g, or rather its Fourier transform, whose energy
should be concentrated around 0. This energy spread is measured
by three parameters.
Implementation
The Fast Windowed Fourier Transform is equivalent to a sequence
of FFTs.
Specrogram examples


The discrete wavelet transform uses a discrete sequence of scales aj
for j<0 with a=21/v, wher V is an integer, called the number of
voices in the octave.
The wavelet support is assumed to be [-K/2,K/2]. For a signal of

size N and 1<=aj<=N/K, a discrete wavelet is defined by sampling
the scale at aj and time (for scale 1) at its integer values, that is
The signal and wavelet are N-periodized. The discrete wavelet

transform of f is
which is a circular convolution between f and ψ1[n]=ψ[-n]. These

circular convolutions are computed with an FFT which requires
O(N log2(N)) operations.
The scalogram is computed from the wavelet transform and a

parobolic interpolation is performed between three succesive scales
aj to better localize the wavelet ridges.
Wavelet transform
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Ondelettes/Fast_Wavelet_Transf.html [11/15/2003 9:43:42 PM]

Wavelet Transform
Wavelet Transform
The wavelet transform replaces the Fourier
transform's sinusoidal waves by a family
generated by translations and dilations of a
window called a wavelet.
It takes two arguments: time and scale.
Outline
The wavelet transform is defined by
where the base atom ψ is a zero average function, centered around

zero with a finite energy. The family of vectors is obtained by
translations and dilatations of the base atom:
The previous function is centered around u, like the windowed

Fourier atom. If the frequency center of ψ is η, then the frequency
center of the dilated function is η/s.
Its time spread is proportional to s. Its frequency spread is

proportional to the inverse of s. Here is an example of
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Ondelettes/Wavelet_Transform.html (1 of 4) [11/15/2003 9:43:47 PM]

Wavelet Transform
Heisenberg boxes of wavelet atoms:
At the finer scales, more Heisenberg boxes can be placed side to

side because there is a better time resolution.
Properties
The wavelet transform has thus a time frequency resolution
which depends on the scale s. Under the condition
it is a complete, stable and redundant representation of the signal;

in particular, the wavelet transform is left invertible. The
redundancy implies the existence of a reproducing kernel.
Scalogram

Wavelet Transform
If η denotes the frequency center of the base wavelet, then the

frequency center of a dilated wavelet is ξ=η/s. The scalogram of a
signal is defined by
The normalized scalogram is .
Choice of Window
As far as the continuous wavelet transform is concerned, a wavelet
is simply a finite energy function with a zero mean. Besides its
Heisenberg box, the most important feature of a wavelet is the
number of its vanishing moments:
The vanishing moments property makes it possible to analyse the

local regularity of a signal.
A theorem caracterizes fast decaying wavelets with n vanishing

moments as the nth derivatives of a fast decaying function.
Implementation
The wavelet transform is computed with a Fast Wavelet Transform.
It computes a discrete transform with circular convolutions, which
are themselves computed with a FFT.
To speed up computations, dyadic wavelets are often used. The

dyadic wavelet transform is implemented by filter banks.

Wavelet Transform
Scalogram examples

Chirps
Chirps are analytic signals which have a particular instantaneous

frequency.
A gaussian chirp is defined by
A linear chirp is defined by
A quadratic chirp is defined by
A hyperbolic chirp is defined by
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Chirps/ChirpsUS.html [11/15/2003 9:43:49 PM]

Gabor Chirps and Wavelets
f is a Gabor chirp if there is such that
A Gabor wavelet is a particular Gabor chirp
with
For
the wavelet is approximatively analytic.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Gabor/Gabor_Wavelet.html [11/15/2003 9:43:53 PM]

Heisenberg Box
Heisenberg Box
The "time frequency localization" of an atom is represented as a
"Heisenberg box" located in the time frequency plane, which is a
rectangle with a time width and a frequency heigth , and time
frequecy center which coincides with the signal's.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/atomes/Heisenberg_box.html [11/15/2003 9:43:55 PM]

Time Frequency Transforms

The windowed Fourier transform replaces the Wavelet Transform
Fourier transform's sinusoidal wave by the The wavelet transform replaces the
product of a sinusoid and a window which is Fourier transform's sinusoidal waves
localized in time. by a family generated by translations
It takes two arguments: time and frequency. and dilations of a window called a
wavelet.
It takes two arguments: time and
Outline
scale.
The windowed Fourier transform is defined by
Outline
The wavelet transform is defined by
It uses an atom which is the product of a sinusoidal wave with a
finite energy symmetric window g. The windowed Fourier
transform family of atoms is obtained by time translations and
frequency modulations of the original window:
where the base atom ψ is a zero average function,
centered around zero with a finite energy. The
family of vectors is obtained by translations and
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Transforms.html (1 of 6) [11/15/2003 9:43:57 PM]

dilatations of the base atom:
This atom has a frequency center ξ and is symmetric with respect

to u.
The time and frequency spreads of these functions are constant.

The family is generated by time and frequency translations of one
The previous function is centered around u, like
atom. Here are examples of Heisenberg boxes of windowed
the windowed Fourier atom. If the frequency
Fourier atoms:
center of ψ is η, then the frequency center of the
dilated function is η/s.
Its time spread is proportional to s. Its frequency

spread is proportional to the inverse of s. Here is
an example of Heisenberg boxes of wavelet
atoms:

Properties
The windowed Fourier transform has a constant time frequency
resolution. This resolution can be can be changed by rescaling the
window g. It is a complete, stable, redundant representation of the
signal. Hence it is invertible. The redundancy implies the existence
of a reproducing kernel.
Spectrogram
The square modulus of the windowed Fourier transform is the
spectrogram of a signal:
Choice of Window
At the finer scales, more Heisenberg boxes can be
The properties of the windowed Fourier transform are determined placed side to side because there is a better time
by the window g, or rather its Fourier transform, whose energy resolution.
should be concentrated around 0. This energy spread is measured
by three parameters. Properties
Implementation The wavelet transform has thus a time frequency

resolution which depends on the scale s. Under
The Fast Windowed Fourier Transform is equivalent to a sequence the condition
of FFTs.

Specrogram examples
it is a complete, stable and redundant

representation of the signal; in particular, the
wavelet transform is left invertible. The
redundancy implies the existence of a reproducing
kernel.
Scalogram
If η denotes the frequency center of the base
wavelet, then the frequency center of a dilated
wavelet is ξ=η/s. The scalogram of a signal is
defined by
The normalized scalogram is .
Choice of Window
As far as the continuous wavelet transform is

concerned, a wavelet is simply a finite energy

function with a zero mean. Besides its Heisenberg
box, the most important feature of a wavelet is the
number of its vanishing moments:
The vanishing moments property makes it

possible to analyse the local regularity of a signal.
A theorem caracterizes fast decaying wavelets

with n vanishing moments as the nth derivatives
of a fast decaying function.
Implementation
The wavelet transform is computed with a Fast
Wavelet Transform. It computes a discrete
transform with circular convolutions, which are
themselves computed with a FFT.
To speed up computations, dyadic wavelets are

often used. The dyadic wavelet transform is
implemented by filter banks.

Scalogram examples

Choice of WFT window
Choice of Fourier
Window
The properties of the windowed Fourier transform are determined
by the window g, or rather its Fourier transform, whose energy
should be concentrated around 0. Three important parameters
evaluate this energy spread:
● the root mean square bandwidth ∆ω defined by
If ∆ω is small, the energy of the window is well concentrated

around 0.
● the maximal amplitude A of its first side lobes, measured in

decibels:
http://cas.ensmp.fr/~chaplais/Wavetour_present...mees/Fourier_Fenetre/Choice_of_WFT_window.html (1 of 2) [11/15/2003 9:43:59 PM]

Choice of WFT window
These side lobes can create "shadows" on each side of the

frequency center.
● The polynomial exponent which describes the decay of the

window's Fourier transform at high frequencies
It sums up the behavior of the Fourier transform beyond the

first side lobes.
The following table gives the values of these parameters for

classical windows, normalized so that g(0) = 1.
Name g(t) ∆ω A p
Rectangle 1 0,89 -13 db 0
Hamming 0,54 + 0,46 cos(2πt) 1,36 -43 db 0
Gaussian exp(-18t2) 1,55 -55 db 0
Hanning cos2(πt) 1,44 -32 db 2
Blackman 0,42 + 0,5 cos(2πt) + 0,08 cos(4πt) 1,68 -58 db 2
http://cas.ensmp.fr/~chaplais/Wavetour_present...mees/Fourier_Fenetre/Choice_of_WFT_window.html (2 of 2) [11/15/2003 9:43:59 PM]

Spectrogram Examples
These few examples are made of analytic signals whose
intanteous frequencies are known.
They illustrate the windowed Fourier transform's ability to localize

instaneous frequencies.
Here is the sum of two parallel linear chirps with its spectrogram.
Now a synthetic signal which is the sum of a linear chirp with an

increasing frequency, a quadratic chirp with a decreasing
frequency, and two modulated gaussians. Below is its spectrogram
and the complex phase of its windowed Fourier transform,
computed with a gaussian window.
The components of this synthetic signal have explicit instantaneous

frequencies.
http://cas.ensmp.fr/~chaplais/Wavetour_present...mees/Fourier_Fenetre/Spectrogram_Examples.html (1 of 3) [11/15/2003 9:44:04 PM]

Here is the sum of two hyperbolic chirps and its spectrogram.

The "instantaneous frequencies" are well tracked in the first

example. On the contrary, the spectrogram loses the frequency of
the hyperbolic chirp when it becomes high too fast. This is due to
the fixed time resolution of the windowed Fourier transform.
In these three examples, it seems that the instantaneous frequencies

is traced by the spectrogram's mawima, provided these frequencies
are not too close.
Windowed Fourier Ridges and Instantaneous Frequencies

Regularity Analysis
Regularity Analysis
The Fourier transform analyses the global regularity of a
function.
The wavelet transform makes it possible to analyze the
pointwise regularity of a function.
A signal is regular if it can be locally approximated by a polynomial. The

definition of the Lipschitz regularity is
Fourier condition
Naturally, this a global regularity condition.
To get conditions on the local or even pointwise regularity of a signal, it is

necessary to use a transform which is localized in time.
Wavelet Transform Condition

Assume that the wavelet has n vanishing moments:
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Regularite/Regularity.html (1 of 3) [11/15/2003 9:44:23 PM]

Regularity Analysis
and has n continuous derivatives with a fast decay.
A fast decaying wavelet has n vanishing moments if and only if its is the nth
derivative of a fast decaying function.
If f is a function which is a little bit more than n times differentiable at point v,

then it can be approximated by a polynomial of degree n. The wavelet
transform of this polynomial is zero; around v, its order is that of the error
between the polynomial and the function. If this error can be uniformly
estimated on an interval, this yields a tool for regularity analysis on an
interval.
This condition relates the pointwise regularity of a signal

to the decay of its wavelet transform's modulus.
It can be extended to an interval and, of course, to the whole real axis.
Example

Regularity Analysis
A signal and its wavelet transform, computed with the derivative of a Gaussian.
Finer scales are at top.
Zero coefficients are represented by a medium gray.
Hence, the regular parts are medium gray.
Notice the cones below the singularities.
Detection of Singularities

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Ondelettes/momentsUS.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Ondelettes/momentsUS.gif [11/15/2003 9:44:25 PM]

Multiresolution Approximation
Are The Foundation of Dyadic
Wavelets
Dyadic wavelets are wavelets which satisfy an
additional scaling property.
This property allows the implementation of a Fast
Dyadic Wavelet Transform with filter banks.
The definition of dyadic wavelets comes from the
definition of multiresolution approximations.
While browsing these pages, you certainly have downloaded

interlaced GIF images. During the download, a progressively
detailed image is displayed on screen.
This idea of consecutive approximations at finer and finer

resolutions is formalized by the concept of multiresolution
approximation (or multiresolution analysis).
Definition
A sequence Vj , j in Z, of subspaces of L2(R) is a multiresolution
approximation if the six following conditions are satisfied:
Condition Interpretation
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Multiresolution_Approx.html (1 of 3) [11/15/2003 9:44:26 PM]

There exists an underlying dyadic

Vj+1 is obtained from Vj by sequence of time grids, e.g., the
a factor 2 rescaling intervals satisfy a geometric
progression with reason 2.
For any j, Vj+1 is a Any low resolution signal is also a
subspace of Vj high resolution signal.
There is an underlying time grid
Vj is 2j translation with step 2j. Condition 1 shows that
invariant. this grid is obtained from the case
j=0 by a 2j rescaling.
The intersection of the Vj is A a zero resolution, the only finite
0 in L2. energy signal is 0.
The union of the Vj is dense At the infinite resolution, all finite

energy signals are perfectly
2
in L . reproduced.
The resolution Vj is generated by a
There is a function such that basis which is obtained by 2j
the integer translations of θ translations of a 2j rescaled θ. A
make a Riesz basis of V0. Riesz basis is a frame of linearly
independent vectors.
A less literary definition is available.
The rescaling of θ does not modify the area of its Heisenberg box,
but it changes the proportions of the box, like for non dyadic
wavelets.
Examples of multiresolution
approximations
What of wavelets?
The wavelets are used to build a basis in which are represented

the details that are gained between a resolution and the next
finer one.
Properties

Algorithme a trous
Fast Dyadic Transform

Algorithme à trous
The fast dyadic wavelet transform is implemented using filter banks. This
implementation is very close to the implementation of the fast (bi)orthogonal
wavelet transform, except that no subsampling is performed.
For any j>=0, let
and the discrete data is likened to the a0[n]. We also define
For a given filter x with coefficients x[n], xj[n] denotes the filter obtained by
inserting 2j-1 zeroes between every x coefficient (hence the French name
"algorithme à trous", which means "holes algorithm"), and let
The algorithme à trous computes the fast dyadic wavelet transform in the
following way:
(the ~ filters are the dual filters of the biorthogonal system).
Compare this algorithm to the decomposition and reconstruction algorithm over a

basis of biorthogonal wavelets. In the decomposition case, the data is convolved
with the symmetrized filter, then the output is subsampled. Here the filter is
"stretched" to take into account the rescaling and the convolution is performed
without any subsampling.
Here is a scheme of the filter bank:
http://cas.ensmp.fr/~chaplais/Wavetour_presenta...ndelettes%20dyadiques/Algorithme_a_trousUS.html (1 of 2) [11/15/2003 9:44:31 PM]

Algorithme a trous
Dyadic Wavelet Transform
Reconstruction From Dyadic Maxima
http://cas.ensmp.fr/~chaplais/Wavetour_presenta...ndelettes%20dyadiques/Algorithme_a_trousUS.html (2 of 2) [11/15/2003 9:44:31 PM]

Perfect Reconstruction Filters
Perfect Reconstruction
and Conjugate Mirror Filter Banks
A perfect reconstruction filter bank decomposes a
signal by filtering and subsampling.
It reconstructs it by inserting zeroes, filtering and
summation.
Definition
A (discrete) two-channel multirate filter bank convolves a signal a0
with a low-pass filter h1[n] = h[-n] and a high-pass filter g1[n] = g[-
n] and then subsamples the output:
a1 [n] = a0 * h1 [2n]
and
d1 [n] = a0* g1 [2n] .
A reconstructed signal a2 is obtained by filtering the zero expanded
signals with a dual low-pass filter h2 and a dual high-pass filter g2.
If z(x) denotes the signal obtained from x by inserting a zero
between every sample, this can be written as:
a2 [n] = z(a1) * h2 [n] + z(d1) * g2 [n] .

The following figure illustrates the decomposition and
reconstruction process.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/Perfect_Reconstruction.html (1 of 2) [11/15/2003 9:44:37 PM]

Perfect Reconstruction Filters
The filter bank is said to be a perfect reconstruction filter bank

when a2 = a0 . If, additionally, h = h2 and g = g2, the filters are
called conjugate mirror filters.
Caracterization
Perfect reconstruction filter banks are caracterized in a theorem by
Vetterli. When the filters have a finite impulse response, the g and
g2 filters can easily be derived from the h and h2 filters, and the
filter synthesis is equivalent to solving
where h and h2 are trigonometric polynomials.
From filters to wavelets
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/Perfect_Reconstruction.html (2 of 2) [11/15/2003 9:44:37 PM]

Normalized Saclogram examples
Normalized Scalogram examples

These few examples are based on the same synthetic signals as
for the windowed Fourier transform.
They illustrate the wavelet transform's ability to localize instaneous

frequencies.
Here is the sum of two parallel linear chirps with its scalogram.
When the frequency increases, the frequency resolution of the
transform decreases.
Here is the normalized scalogram and the complex phase of the

wavelet transform of a synthetic signal which is the sum of a linear
chirp with an increasing frequency, a quadratic chirp with a
decreasing frequency, and two modulated gaussians. Their
computations have been performed with a Gabor wavelet.
The makup of this signal explicitly introduces instantaneous

frequencies.
http://cas.ensmp.fr/~chaplais/Wavetour_presenta...transformees/Ondelettes/Scalogram_Examples.html (1 of 3) [11/15/2003 9:44:52 PM]

Here is the sum of two hyperbolic chirps and its scalogram.
In the first example, the instantaneous frequencies are more blurred

in the scalogram than in the spectrogram. On the other hand, the
variable time resolution of the wavelet transform makes it possible
to track the hyperbolic frequency across time. The decreasing
frequency resolution is masked by the vertical asymptotic
tendency.

Wavelet Ridges and Instantaneous Frequencies

Frames
Frames
Frames are a stable, possibly redundant,
representation of signals.
It is a generalization of the concept of basis in a
linear space.
A frame is a family of vectors which can represent any finite

energy signal by the sequence of its inner products with the vectors
of the family. However, it may be possible that not all sequences of
reals represent an finite energy signal. Oversampling is an example
of a represention of signals in a frame. One can see that not all
sequences of values may represent a sequence of samples. In
general, frames are a stable and redundant representation of
signals.
Definition
A family of vectors in a Hilbert space H is a frame of H if
there are two constants A>0 and B>0 such that, for any f in H,
If A=B, the frame is said to be tight.
A Riesz basis is a frame whose vectors are linearly independant.
Example: consider a family of three vectors in the plane which are obtained by
succesive rotations of a third of turn of one vector. This family is a tight frame of the
plane, with A=B=3/2.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/frames/Frames_IntroUS.html (1 of 3) [11/15/2003 9:44:57 PM]

Frames
Properties
The frame vectors are supposed to be of unit norm.
If the frame vectors are independent, then A<=1<=B. The frame is

then an orthonormal basis if and only if A=B=1. If A>1, then the
frame is redundant. A finite family is always a frame of the linear
space that it generates.
Pseudo Inverse
U denotes the operateur which transforms a signal f into the
sequence of its frame inner products.
U has one or an infinity of left inverses.
The pseudo-inverse of U is the left inverse of U which is zero

on the orthogonal complement of the image of U. It is the
minimum norm left inverse.
It is used to build a signal approximation from any sequence of real

numbers. The computation of (U*U)-1f can be performed by a
conjugate gradient algorithm.
Dual Frame
The image of the frame through (U*U)-1 is a frame called
the dual frame. For any f in H,
and

Frames
If the original frame is a Riesz basis, then the two frames form a
biorthogonal basis system, that is
Windowed Fourier Frames and Wavelet Frames

Windowed Frames
Windowed Fourier Wavelet Frames

Frames To cover the time frequency plane with wavelet
Heisenberg boxes, a regular grid is not used; time
The Heisenberg boxes of windowed
steps which are inverse proportional to the
Fourier atoms have dimensions which do frequecy step are used instead, the latter being
not depend on their time or frequency itself proportional to the scale.
center. To get a frame after discretization
of the windowed Fourier transform, a
tiling of the time frequency plane by the
discrete family of Heisenberg boxes is
desirable. Hence, it only logical to use a
regular rectangular grid to place the time
frequency centers.
The wavelet is assumed to satisfy the

reconstruction condition
which garantees the invertibility of the wavelet

Daubechies gives the following necessary
transform. Daubechies gives necessary conditions
condition on the tiling to yield a frame:
for the previous tiling to yield a frame:
Sufficient conditions also exist.
Two important results should be

emphasized:
● there is no compactly supported, Sufficient conditions also exist.

continuously differentiable window
that generates an orthogonal The following differences with the windowed
windowed Fourier basis (Balian- Fourier frames should be emphasized:
Low theorem).
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/frames/Windowed_Frames.html (1 of 3) [11/15/2003 9:45:01 PM]

Windowed Frames
● the dual frame of a windowed ● there are continuously differentable

Fourier frame is also a window wavelets that generate frames (look at the
Fourier frame. construction of wavelet bases for more)
● in the general case, the dual frame of a
wavelet frame is not a wavelet frame.
However, in the cases of bases, a dual
wavelet basis can be built by other means
(look at the wavelet bases for more,
especially the biorthogonal ones)
Translation invariance
In both cases (Fourier or wavelets), the frame
representation has the drawback of not being
translation invariant with respect to time or
frequency. Now, most interesting signal patterns
are not naturally synchronized with frame
intervals. In particular, the structure of a signal
may be degraded at the lower resolutions.
This motivates the study of the dyadic wavelet

transform, which is discrete in scale but not in
time (in practice, this means that signals are
oversampled when switching to coarser
resolutions).
Another time invariant represenation is the

representation by dyadic wavelet maxima. It is
less redundant, but is not complete.
Why wavelet bases are studied

nonetheless
In practice, the dyadic wavelet transform is
implemented by perfect reconstruction filter
banks. These fast filter banks correspond to
wavelet bases which are built from
multiresolution approximations.

Windowed Frames
Multiresolution Approximations and Wavelet

Bases

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier_Fenetre/window-design.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier_Fenetre/window-design.gif [11/15/2003 9:45:03 PM]

Windowed Fourier Rigdes
Windowed Fourier Ridges

The windowed Fourier ridges are the maxima
points of the spectrogram.
They indicate the instantaneous frequencies within
the limits of the transform's resolution.
The latter is determined by the Heisenberg boxes
which tile the time frequency plane.
Windows Used
The windows g used here are symmetric with respect to 0 and have
a support within [-1/2,1/2], as in the previous table.
The windowed Fourier ridges are the maxima points of the

spectrogram. If the amplitude and frequency have a small variation
within the Fourier window, and if the instantaneous frequency is
higher than the window's passing band, then the frequencies
which maximize the spectrogram approximate the
instantaneous frequencies. At these points, the complex phase of
the transform is almost constant.
Time Frequency Resolution

The windowed Fourier ridges of the sum of two analytic signals
can discriminate their two instantaneous frequencies if their
difference is greater than the scaled window's bandwidth:
http://cas.ensmp.fr/~chaplais/Wavetour_presentat...transformees/Fourier_Fenetre/Fourier_Ridges.html (1 of 2) [11/15/2003 9:45:09 PM]

Windowed Fourier Rigdes
where s is the scaling which has been applied to the Fourier

window, and ∆ω is the bandwidth of the unscaled window g.
This is a condition on the absolute frequency difference. It is

related to the structure of the time frequency tiling.
Hence, the windowed Fourier ridges can detect instantaneous

frequencies provided they are not too close.
Examples of Windowed Fourier Ridges
http://cas.ensmp.fr/~chaplais/Wavetour_presentat...transformees/Fourier_Fenetre/Fourier_Ridges.html (2 of 2) [11/15/2003 9:45:09 PM]

Wavelets
Orthogonal Wavelets
Let us recall that a multiresolution approximation is a nested Biorthogonal Wavelets
sequence of linear spaces. The orthogonal complement Wj of Vj in
Vj-1 can be thus defined: Biorthogonal wavelets are defined similarly to
orthogonal wavelets, except that the starting point is
biorthogonal multiresolution approximations. The
following decompositions are performed:
Then there is a function ψ such that the family ψj,n(t) = 2-j/2 ψ(2-jt-
n), n in Z, is an orthonormal basis of Wj. The family ψj,n, j in Z
and n in Z, is an orthonormal basis of L2 and
Like in the orthogonal case, a signal in L2 can be

written as
ψ is an orthogonal wavelet associated to the multiresolution
approximation. A signal f in L2 can be decomposed as
where φ is an orthogonal scaling function of the multiresolution.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/Wavelets.html (1 of 4) [11/15/2003 9:45:17 PM]

Wavelets
A theorem by Mallat and Meyer builds an orthogonal wavelet from

a scaling function.
Example
Here is a cubic spline scaling function and the corresponding cubic
spline Battle-Lemarié wavelet, and their Fourier transform. The
wavelet is a cubic spline because it is a linear combination of cubic
splines.
Example
Below is a biorthogonal system which includes a cubic
B-spline. Dropping the orthogonality constraint makes
possible to have both regularity and symmetry.
The wavelet is not compactly supported.
From the real axis to the interval

Wavelets
Wavelet bases are bases of L2(R). There are several methods to

transform them into wavelet bases over an interval. Once
discretized, they are used to process finite signals.
From dimension 1 to dimension 2

There are several methods to build wavelet bases on functional
spaces in dimension greater than 1. The simplest ones uses
separable wavelets.
Biorthogonal cubic B-spline scaling Dual scaling function

function
Properties
Biorthogonal spline wavelet Dual Wavelet
The construction of biorthogonal wavelets over the

interval or in dimension 2 will not be presented here. It
follows the same lines as in the orthogonal case.

Wavelets
Properties


Dyadic wavelet transforms are scale samples of
wavelet transforms following a geometric
sequence of ratio 2. Time is not sampled.
This transform uses dyadic wavelets.
It is implemented by perfect reconstruction filter
banks.
Definition
The dyadic wavelet transform of f is defined by
It defines a stable complete representation if its Heisenberg boxes

cover all of the frequency axis, that is, if there exist A et B such
that
The family of dyadic wavelets is a frame of L2(R).
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes%20dyadiques/Dyadic_Transform.html (1 of 2) [11/15/2003 9:45:21 PM]

Wavelet synthesis
To build dyadic wavelets, it is sufficient to satisfy the previous
condition. To do so, it is possible to proceed as for the construction
of orthogonal and biorthogonal wavelet bases, using conjugate
mirror or perfect reconstruction filter banks.
The wavelets satisfy then scaling equations and the fast dyadic
wavelet transform is implemented using filter banks.
Implementation
The fast dyadic wavelet transform uses the same filters as for the
computation of the fast wavelet transform of a discrete signal,
except that no subsampling is performed.
Back to top
Next path
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes%20dyadiques/Dyadic_Transform.html (2 of 2) [11/15/2003 9:45:21 PM]

Reconstruction From Dyadic

Wavelet Modulus Maxima
A signal is representated by its low pass
approximation and the modulus maxima of its
dyadic wavelet transform.
This representation allows an almost perfect
reconstruction of a signal.
Outline
The continuous wavelet transform detects isolated singularities
with their order of singularity. The regular part of the signal is
coded in its coarsest approximation. It is sensible to try to
reconstruct a signal from this coarse resolution and from its
wavelet modulus maxima.
In practice, only the dyadic wavelet transform is considered to take

advantage of the fast algorithme à trous which implemented by
filter banks.
From a theoretical point of view, Meyer and Berman have proved

that the representation by dyadic maxima is not complete because
several signals may exhibit the same wavelet maxima.
In practice, numerical experiments have shown that it is possible to

reconstruct usual signals with a relative mean sqaure error smaller
than 10-2. On images, the difference is not visible.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes%20dyadiques/Dyadic_Maxima.html (1 of 2) [11/15/2003 9:45:23 PM]

Implementation
A signal is to be reconstructed from the values and locations uj,p of
its wavelet modulus maxima, j being the scale and p the time
localization. This difficult problem is replaced in practice by a
simpler one which consists in finding a minimum norm signal
among those which have the assigned wavelet coefficients at the
maxima locations. Solving this problem tends to create signal with
modulus maxima at the right locations with the correct values.
Since this problem actually bears on discrete signals, this

simplified probleme is an inverse frame problem, which can be
solved using a conjugate gradient algorithm. To this reconstruction
a previously stored low frequency component defined by the
sample averages is added.
An example in PDF format (32 Kb) is available. Here is a preview

of it:
Images, Edge Dectection and Reconstruction
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes%20dyadiques/Dyadic_Maxima.html (2 of 2) [11/15/2003 9:45:23 PM]

Detection of singularities
Wavelet transform modulus maxima are related to the
singularities of the signal.
More precisely, the following theorem proves that there cannot be a

singularity without a local maximum of the wavelet transform at the finer
scales.
This theorem indicates the presence of a maximum at the finer scales where a
singularity occurs. In the general case, is sequence of modulus maxima is
detected which converges to the singularity. Below are the modulus maxima of
the previous example:
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Regularite/Detection_of_singularities.html (1 of 3) [11/15/2003 9:45:51 PM]

Modulus maxima are in yellow (light grey on greyscale monitor).

A PDF file shows the connection between wavelet modulus maxima and the signal singularities.
Since log2(s) >=0 because of the discretization, the detection on the wavelet transform
is restricted to log2(s)>=1 to preserve the continuous case approximation.
Warning: these are the modulus maxima of the wavelet transform.

Instantanuous frequencies are detected from the maxima of the normalized
scalogram:
which differs in two ways: normalization, and the fact that the
variable is homogeneous to a frequency, and not to a scale.
When the wavelet is the nth derivative of a gaussian, the maxima curves are
connected and go through all of the finer scales.
The decay rate of the maxima along the curves indicate the order of the isolated
singularities (this a consequence of theorems 6.4 et 6.6 when extended to an
interval):

The modulus maxima are displayed as a function of the scale in log-log axes,
and the slope gives the estimated singularity order. Below is such a curve for
two singularities: the solid line corresponds to the singularity at t=14 and the
dotted line to the singularity at t=108. Fine scales are on the left.
For t=14, the slope is 1/2, and the signal is 0-Lipschitz here, that is, it has a
discontinuity. For t=108, the slope is close to 1, which indicates that the signal
is 1/2 Lipschitz here.

Biorthogonal Wavelets and Discrete Filters
Biorthogonal Wavelets
and Discrete Filters
The scaling equations on the scaling functions and wavelets show
that the decomposition and reconstruction of a signal from a
resolution to the next one is implemented by perfect reconstruction
filter banks.
The scaling equations imply the coefficients a1 [n] and d1 [n] of a

signal in Vj and Wj are computed from its coefficients a0[n] in Vj-1
by applying the filters h and g and subsampling the output:
a1 [n] = a0 * h1 [2n]
and
d1 [n] = a0* g1 [2n] .
with h1[n] = h[-n] and g1[n] = g[-n].
In practice this recursion is initialized by considering that the

discrete signal samples are some fine resolution coefficients.
The coefficients of h and g are defined by the scaling equations
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/Biortho_Wave_and_filt.html (1 of 3) [11/15/2003 9:45:56 PM]

or, in the Fourier domain:
Conversely, a0[n] is reconstructed from a1 [n] and d1 [n] by

inserting zeroes between two consecutive samples and summing
their convolutions with the dual filters h2 et g2 which define the
dual scaling equations:
a0 [n] = z(a1) * h2 [n] + z(d1) * g2 [n]

where the z operator represents the insertion of zeroes.
The coefficients of h2 and g2 are defined by the scaling equations

This algorithm is used to evaluate the scaling functions and

wavelets. Indeed the coefficients of a scaling function (resp.
wavelet) are all zero but for one within their on resolution (resp.
detail) space basis. The reconstruction algorithm provides the
coefficients in the finer resolutions. For high resolutions, the
scaling coefficients are considered to be samples of the function.
Hence the construction of biorthogonal

wavelets is equivalent to the synthesis
of perfect reconstruction filters having a
stability property.
Filtering

Vetterli
g and g2 can be eliminated from the previous equations, which

leads to the necessary condition:
For filite impulse response filters, the Fourier transforms of such

filters are trigonometric polynomials, and conditions (7.121) and
(7.129) can be interpreted as Bezout identities in the ring of
trigonometric polynomials. In this ring, units are trigonometric
monomials. Equations (7.121) - (7.122) form a linear system with
respect to h2 and g2, and it can be shown that the associated matrix
is unimodular, i.e., its determinant is a trigonometric monomial.
then there exists a real number a and an integer l such that
Back to perfect reconstruction filters
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/VitterliUS.html [11/15/2003 9:45:59 PM]


Wavelets and scaling functions
Biorthogonal wavelets and scaling functions are caracterized by a perfect
reconstruction filter bank; orthogonal wavelets and scaling functions are
caracterized by a pair of conjugate mirror filters. Nonetheless, a perfect
reconstruction filter bank (or any pair of conjugate mirror filters) does not
necessarily generate a wavelet system. Indeed, some attention has to be paid
to the stability of the decomposition and reconstruction schemes as the
number of scales increases, that is when the number of filter bank cascades
goes to the infinity. This is expressed by an additional condition (7.37) on the
conjugate mirror filter h for it to define a scaling function.
Perfect reconstruction filter banks and

algorithme à trous
The decomposition can be performed on the signal a1 to generate a signal a2
and a signal d2; repeating this construction produces a low resolution signal
aj and a sequence of detail signals d1 .... dj.
A recursive decomposition which similar to the previous one can be

performed by the algorithme à trous to generate low resolution signal Aj and
a sequence of detail signals D1 .... Dj. The two decompositions are related by
the following equations:
aj[n] = Aj[2jn]
dj[n] = Dj[2jn]
http://cas.ensmp.fr/~chaplais/Wavetour_presenta...ndelettes%20dyadiques/FromFilters2Wavelets.html (1 of 2) [11/15/2003 9:46:03 PM]

From algorithme à trous to scaling functions

In the Fourier domain, the transfer between a0 and Aj is
Let us operate a time rescaling T = 2-jt so that the interval between the non
zero coefficients of the slower filter is always one. Then the interval between
the non zero coefficients of the tightest filter is 2-j. The transfer becomes
Let j go to the infinity. If the previous transfer converges in L2, then its limit
is the Fourier transform of a finite energy signal which necessarily satisfies a
scaling equation:
Such functions are at the core of multiresolution analysis, which is itself the
sarting point for the definition of dyadic wavelets.
Multiresolution analysis
Filter Synthesis
http://cas.ensmp.fr/~chaplais/Wavetour_presenta...ndelettes%20dyadiques/FromFilters2Wavelets.html (2 of 2) [11/15/2003 9:46:03 PM]

Wavelet Ridges
Wavelet Ridges
The wavelet ridges are the maxima points of the
normalized scalogram.
They indicate the instantaneous frequencies within
the limits of the transform's resolution.
The latter is determined by the Heisenberg boxes
which tile the time frequency plane.
Wavelets used
Approximatively analytic wavelets are used:
like Gabor wavelets. The atoms are similar to a windowed Fourier

transform's, but, after rescaling, the window width is proportional
to the "frequency" ξ=η/s.
Hence, similar windows are used, but with a different time

frequency tiling.
The wavelet ridges are the maxima points of the normalized

scalogram. Under conditions which are similar to the
spectrogram's, the frequencies which maximize the normalized
scalogram approximate the instantaneous frequencies. The
difference is that the time frequency resolution structure is
different.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Ondelettes/Wavelet_Ridges.html (1 of 2) [11/15/2003 9:46:09 PM]

Wavelet Ridges
Time frequency resolution

The wavelet ridges of the sum of two analytic signals can
discriminate their two instantaneous frequencies if their relative
differences are greater than the relative wavelet bandwidth:
and
where ∆ω is the wavelet bandwidth and η its frequency center.
These conditions bear on the relative frequency differences. They

are related to the structure of the time frequency tiling.
Hence, the wavelet ridges can detect instantaneous frequencies

provided their relative distances are not too small.
Examples of Wavelet Ridges
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Ondelettes/Wavelet_Ridges.html (2 of 2) [11/15/2003 9:46:09 PM]

Orthogonal Multiresolution Approximations
Orthogonal
Orthogonalization
The Riesz basis can be transformed into an orthogonal basis generated by
integer translations of an elementary function, which is called a scaling
function. It is a particular case of a biorthogonal system where both bases
are equal.
Scaling function built from cubic spline approximations and its Fourier transform.
Observe the time frequency localization.
The scaling function is a cubic spline because it is generated by cubic splines..
Scaling Equation
One can verify that the other resolutions are generated by a suitable
dilatation of these bases of translated atoms. Since the resolutions are
embedded, there is necessarily a sequence of real numbers h[n] such that
or, in the Fourier domain
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Orthogonal_Multires.html (1 of 3) [11/15/2003 9:46:16 PM]

assuming the Fourier transform is continuous.
It is proved that h is a conjugate mirror filter. It entirely determines the

scaling function and most of its properties. In particular, the scaling
function is compactly supported if and only if h has a finite number of
non zero coefficients. h is sais to be a Finite Impulse Response (FIR)
filter. For more info, see the page on the properties of of orthogonal
wavelets and how they are related the the filter h.
As an example, the filter coefficients that correspond to the cubic spline

scaling function are given.
The study of the filters h which generate multiresolution approximations

has produced many important theorems. The following one gives
necessary and sufficient conditions for h to generate a scaling function:
Condition (7.35) means that h is a conjugate mirror filter. Condition

(7.36) is simply a normalization. Condition (7.37) garantees that the
function defined by (7.38) has a finte energy.

The scaling equations shows that the scaling coefficients a1 [n] = < a ,
ϕ(t/2-n) > of a in V1 are obtained from the scaling coefficients a0 [n] = <
a , ϕ(t-n) > in V0 by a convolution with the conjugate mirror filter h and a
subsampling:
a1 [n] = a0 * h1 [2n]
Back to Multiresolution Approximations

Properties of Orthogonal Wavelets
Properties
Dilated wavelets are related by a scaling equation.
Rescaling can be interpreted as discrete filtering.
Vanishing moments, support, regularity and
symmetry of the wavelet and scaling function are
determined by the scaling filter.
Scaling equation
ψ(t/2) and φ(t) are related by a scaling equation, similar to the
equation which relates ϕ(t/2) and ϕ(t). It is a consequence of
(7.60):
In the Fourier domain, this equation becomes
The h and g filters are a conjugate mirror filter bank.
Vanishing moments
A wavelet has m vanishing moments if and only if its scaling
function can generate polynomials of degree smaller than or equal
to m. While this property is used to describe the approximating
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/Properties_of_Orth_Wave.html (1 of 4) [11/15/2003 9:46:21 PM]
power of scaling functions, in the wavelet case it has a "dual"

usage, e.g. the possibility to caracterize the order of isolated
singularities.
The number of vanishing moments is entirely determined by the

coefficients h[n] of the filter h which is featured in the scaling
equation.
If the Fourier transform of the wavelet is p continuously

differentiable, then the three following conditions are equivalent:
● the wavelet ψ has p vanishing moments

● the scaling function ϕ can generate
polynomials of degree smaller than or equal to
p
● the transfer function of the filter h and its p-1
first derivatives vanish at ω=π.
Compact support
Compactly supported wavelets and scaling functions exist.
The scaling function is compactly supported if and only if the filter

h has a finite support, and their supports are the same. If the
support of the scaling function is [N1,N2], then the wavelet support
is [(N1-N2+1)/2,(N2-N1+1)2].
Atoms are thus compactly supported if and only if

the filter h is.
Daubechies has proved that, to generate an orthogonal wavelet with
p vanishing moment, a filter h with minimum length 2p had to be

used. Daubechies filters, which generate Daubechies wavelets,

have a length of 2p. The Daubechies filter coefficients are available
as ASCII text files which can be used in a spreadsheet, for instance.
Regularity
Wavelet regularity is much less important than their vanishing
moments. Il is studied in a theorem by Tchamitchian
The following two properties are important:
● there is no compactly supported orthogonal wavelet which

indefinitely differentiable
● for Daubechies wavelets with a large p, the scaling function
and wavelet are l-Lipschitz, where l is of the order of 0.2 p.
For large classes of orthogonal wavelets, more regularity
implies more vanishing moments.
Meyer wavelets are indefinitely differentiable orthogonal wavelets,

with an infinite support. They are generally implemented in the
Fourier domain.
Symmetry
Symmetric scaling functions and wavelets are important because
they are used to build bases of regular wavelets over an interval,
rather than the real axis. Daubechies has proved that, for a wavelet
to be symmetric or antisymmetric, its filter must have a linear
complex phase, and the only symmetric compactly supported
conjugate mirror filter is the Haar filter, which corresponds to a
discontinuous wavelet with one vanishing moment. Besides the
Haar wavelet, there is no symmetric compactly supported
orthogonal wavelet.

Orthogonal Wavelets and Discrete Filters

Coefficients du filtre spline cubique
n h[n]
0 0,766130398
1,-1 0,433923147
2,-2 -0,050201753
3,-3 -0,110036987
4,-4 0,032080869
5,-5 0,042068328
6,-6 -0,017176331
7,-7 -0,017982291
8,-8 0,008685294
9,-9 0,008201477
10,-10 -0,004353840
11,-11 -0,003882426
12,-12 0,002186714
13,-13 0,001882120
14,-14 -0,001103748
15,-15 -0,000927187
16,-16 0,000559952
17,-17 0,000462093
18,-18 -0,000285414
19,-19 -0,000232304
20,-20 0,000146098
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Coefficients_du_filtre_spl.html [11/15/2003 9:46:23 PM]

Properties of Multiresolution Approximations
Properties
Orhogonality and biorthogonality
When the Riesz basis is an orthogonal basis, the multiresolution

approximation is orthogonal, and the base atom is called a scaling
function. It is always possible to orthogonalize a multiresolution
approximation.
However, orthogonalities imposes some constraints that may not be

desirable. One of the most important is that a compactly supported
(orthogonal) scaling function cannot symmetric and continuous.
The symmetry is useful in the analysis of finite signals.
Some of these restrictions (notably the absence of symmetry) can

be avoided by using biorthogonal multiresolution approximations.
Approximation
The "approximation" denomination means that an orthogonal or

biorthogonal multiresolution analysis (or approximation) can be
related to a sequence of respectively orthogonal or oblique
projectors, which efficiently approximate regular signals. The order
of approximation is determined by the degree of the polynomials
that can be reconstructed in the resolutions.
Construction and digital filters
Multiresolution approximations are determined by one or two

atoms φ which generate respectively the orthogonal or biorthogonal
multiresolutions.
By definition of a multiresolution, φ(t/2) is a linear combination of

the φ(t-n). This relation is called a scaling equation:
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Multiresolution_Properties.html (1 of 3) [11/15/2003 9:46:27 PM]

In the Fourier domain, this becomes
The coefficients h[n] in the scaling equation entirely determine φ,

and finding them is equivalent to the design of a filter bank, plus
some stability conditions to be able to generate L2.
Wavelets
As the scale j gets finer the approximations becomes more accurate

(see Lena). Switching from the resolution j to j-1 adds details to the
signal. The same way that approximations can be decomposed on
resolution bases, these extra details can be decomposed in detail
bases.
Details bases, like resolution bases, ore obtained by translating a

single resealed atom. This atom is called a wavelet. The order of
approximation of the multiresolution is equal to the number of
vanishing moments of the wavelet. It also represents the wavelet's
ability to detect the isolated singularities of a signal.
Wavelet Bases


http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/MallatMeyer2US.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/MallatMeyer2US.gif [11/15/2003 9:46:32 PM]

Wavelets Over an Interval

Until now, only wavelets over the real axis have been considered,
e.g. wavelets that are suited to the analysis of signals defined over
the whole axis. In most cases, signals are compactly supported;
images, in particular, are explicitely defined over a rectangle of
pixels.
The wavelets considered here are compactly supported.
A [0,N] supported signal can be represented as the product of a

general signal with the caracteristic function of [0,N]. The
discontinuities of this function require special attention. Three
methods are known to handle them, the last one being the most
efficient.
Wavelet periodization
The wavelets are periodized by the following transformation:
with j<=log2N. This is equivalent to a signal periodization.
Wavelets which are completely inside the interval are not changed.
Wavelets that ovelap the boudaries are cut into two pieces loacated
at the left and right edges of the interval. Separately, each of the
pieces have no vanishing moment. This creates large wavelet
coefficients when the periodized signal is not itself continuous.
Wavelet folding
To bypass this problem, the signal is symmetrically folded around
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/Interval_Wavelets.html (1 of 3) [11/15/2003 9:46:40 PM]

the right edge of the interval and periodized over the double sized
interval.
This yields a continuous periodic signal.
Porting the signal transformation to the wavelet basis shows that

the vector family is a basis of L2[0,N] if the wavelet is symmetric
or antisymmetric. This puts orthogonal bases asides.
In fact, the continuity problem reappears at the next derivative. The

following approach takes the problem at the root, which is how to
make wavelets over an interval with vanishing moments.
Edge wavelets
Boundary effects are explicitely handled. Consider an Daubechies
orthogonal basis with p vanishing moments.
From the Strang et Fix conditions, it appears that there exists a

polynomial θk of degree k such that:
for k<p.
This equation is multiplied by the caracteristic function of [0,N].

Assuming that the support of ϕ is [-p+1,p], scaling functions with
indices p<=k<N-p are not changed by this restriction. To recover
the Strang and Fix condition on the interval, p "left" edge scaling
function and p "right" edge scaling functions are to be found such

that
If this equation is satisfied, it reamains valid after rescaling since

the nk, up to a power of 2, are the scaling coefficients of θk at all
resolutions. There remains to find the filters h and H which satisfy
the scaling equation:
where denotes the whole set of scaling functions obtained by

translation at the resolution j, and to verify the orthogonality
condition.
The coefficients of these filters are available in Wavelab, using the

function MakeCDJVFilter.
Back to Wavelet Bases

Strang and Fix conditions
What approximations?
An orthogonal multiresolution approximation defines an
orthogonal projector on each of the resolution spaces. In the
biorthogonal case, the decomposition
defines a (non necessarily orthogonal) projector on V0, and, after

rescaling, a projector on each resolution Vj. The projection of a
signal f is:
This projection is an approximation of f under the following

conditions:
The sufficient condition can be interpreted as follows: the

projection on Vj is able to "catch" Taylor expansions of f up to
degree p at intervals of length 2j.
The general Stang and Fix conditions are available with proof
(PDF v.3, 107 K).
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Strang_and_Fix.html (1 of 2) [11/15/2003 9:46:53 PM]

Strang and Fix conditions
Remark
The definition of multiresolution approximations specifies
Is this compatible with the previous theorem? (Answer)
Back to multiresolution approximations
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Strang_and_Fix.html (2 of 2) [11/15/2003 9:46:53 PM]

MakeCDJVFilter
MakeCDJVFilter -- Set up
filters for CDJV Wavelet
Transform
Usage
[a,b,c] = MakeCDJVFilter(request,degree)
Inputs
request
string: 'HighPass', 'LowPass', 'Precondition',

'Postcondition'
degree
integer: 2 or 3 (number of vanishing moments)
Outputs
a,b,c filter, left edge filter, right edge
filter
('HighPass', 'LowPass')
a conditioning matrix ('Precondition',

'Postcondition')
Description
CDJV have developed an algorithm for wavelets on
the interval which preserves the orthogonality,
vanishing moments, smoothness, and compact
support of Daubechies wavelets on the line.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/MakeCDJVFilter.html (1 of 2) [11/15/2003 9:46:57 PM]

MakeCDJVFilter
The algorithm for wavelets on the interval of

CDJV involves four objects not present in the
usual periodized algorithm: right edge filters,
left edge filters, and pre- and post-
conditioning operators.
These objects are supplied by appropriate

requests to MakeCDJVFilter.
See Also
IWT_CDJV, FWT_CDJV, CDJVDyadDown
References
Cohen, Daubechies, Jawerth and Vial, 1992.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/MakeCDJVFilter.html (2 of 2) [11/15/2003 9:46:57 PM]

Two Dimensional Wavelets
Two Dimensional Wavelets

The simplest way to build two dimensional wavelet bases is to use
separable products on a one dimensional wavelet and scaling
function. This yields the following scaling function
and there are three wavelets:
Back to Wavelets Bases
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/Wavelets_in_dim_2.html [11/15/2003 9:47:00 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/ScalingEq.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/ScalingEq.gif [11/15/2003 9:47:01 PM]

Index of /~chaplais/Wavetour_presentation/Tables/Daubechies
Index of
/~chaplais/Wavetour_presentation/Tables/Daubechies
Name Last modified Size Description
Parent Directory -
Daubechies_coefficients 04-Jun-1999 12:56 2.6K
Daubechies_coefficie..> 04-Jun-1999 12:56 2.7K
Apache/2.0.43 Server at cas.ensmp.fr Port 80
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/ [11/15/2003 9:47:02 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/TchamitchianUS.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/TchamitchianUS.gif [11/15/2003 9:47:04 PM]

Orthogonal Wavelets
and Discrete Filters
g and h are conjugate mirror filters.
Conjugate mirror filters are a particular instance of perfect
reconstruction filter banks. The dyadic nature of multiresolution
approximations are closely related to the possibility of
implementing elementary signal subsampling by erasing one
sample every two, and elementary oversampling by zero insertion
between two consecutive samples.
The coefficients 1 [n] and d1 [n] of a signal in Vj and Wj are

computed from its coefficients a0[n] in Vj-1 by applying conjugate
mirror filters and subsampling the output:
a1 [n] = a0 * h1 [2n]
and
d1 [n] = a0* g1 [2n] .

The coefficients of h are defined by the scaling equation
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/Orth_Wavelets_and_Filters.html (1 of 3) [11/15/2003 9:47:08 PM]

and the coefficients of g are defined by the wavelet scaling

equation
Conversely, a0[n] is reconstructed from a1 [n] and d1 [n] by

inserting zeroes between two consecutive samples and summing
their convolutions with h and g:
a0 [n] = z(a1) * h [n] + z(d1) * g [n]

Wavelets and scaling functions are evaluated as in the orthogonal

case.
Hence the construction of orthogonal

wavelets is equivalent to the synthesis
of conjugate mirror filters having a
stability property.
We will concentrate on Finite Impulse Response
filters, e.g., to compactly supported wavelets.
Filtering

Properties of Biorthogonal Wavelets
Properties
Properties of biorthogonal wavelets are to be
compared to the orthogonal case.
Scaling equation
As in the orthogonal case, ψ(t) and ϕ(t/2) are related by a scaling
equation which is a consequence of the inclusions of the resolution
spaces from coarse to fine:
Similar equations exist for the dual functions which determine the
filters h2 and g2.
Vanishing moments
A biorthogonal wavelet has m vanishing moments if and only if its
dual scaling function generates polynomials up to degree m. This
can be verified by looking at the biorthogonal decomposition
formulas.
Hence there is an equivalence theorem between vanishing moments

and the number of zeroes of the filter's transfer, provided that
duality has to be taken into account. Thus the following three
properties are equivalent:
● the wavelet ψ has p vanishing moments

● the dual scaling function ϕ2 generates
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/Properties_of_Biorth_Wave.html (1 of 4) [11/15/2003 9:47:11 PM]
polynomials up to degree p
● the transfer function of the dual filter h2 and it
p-1 first derivatives vanish at ω=π
and the dual result is also valid. Duality appears naturally, because
the filters determine the degree of the polynomials which can be
generated by the scaling function, and this degree is equal to the
number of vanishing moments of the dual wavelet.
Compact support
If the filters h et h2 have a finite support, then the scaling functions
have the same support, and the wavelets are compactly supported.
If the supports of the scaling functions are respectively [N1,N2]
and [M1,M2], then the corresponding wavelets have support [(N1-
M2+1)/2,(N2-M1+1)/2] and [(M1-N2+1)/2,(M2-N1+1)]
respectively.
The atoms are thus compactly supported if and

only if the filters h et h2 are.
Regularity
Tchamitchian's theorem provides again a sufficient regularity
condition. Remember that this condition bears on the filter h which
determines the scaling equation. Hence the regularity of the primal
atoms are related to the primal filters.
Wavelet balancing
Consider the following decomposition of f:

The number of vanishing moments of a wavelet is determined by

its dual filter. It corresponds to the approximating power of the
dual multiresolution sequence. This is why it is preferred to
synthesize a decomposition filter h with many vanishing moments,
and possibly with a small support.
On the other hand, this same filter h determines the regularity of ϕ,

and hence of ψ. This regularity increases with the number of
vanishing moments, that is, with the number of zeroes of h.
Symmetry
Unlike the orthogonal case, it is possible to synthesize biorthogonal
wavelets and scaling functions which are symmetric or
antisymmetric and compactly supported. This makes it possible to
use the folding technique to build wavelets on an interval.
If the filters h and h2 have and odd length and are symmetric with
respect to 0, then the scaling functions have an even length and are
symmetric, and the wavelets are also symmetric. If the filters have
an even length and are symmetric with respect to n=1/2, then the
scaling functions are symmetric with respect to n=1/2, while the
wavelets are antisymmetric.
Example
Spline wavelets and scaling functions are an interesting example of
biorthogonal systems. One of the scaling functions is a B-spline. A
coefficient table is available. There is a general closed form
formula for these filters.


http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/decompjBiortho.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/decompjBiortho.gif [11/15/2003 9:47:14 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Splines_biortho/splines
n p ,\tilde p h[n] {\tilde h}[n]

0 p = 2 , \tilde p =4 0,70710678119 0,99436891104
1,-1 0,35355339059 0,41984465133
2,-2 -0,1767766953
3,-3 -0,06629126074
4,-4 0,03314563037
0,1 p= 3 , \tilde p=7 0,53033008589 0,9516421219
-1,2 0,1767766953 -0,02649924095
-2,3 -0,30115912592
-3,4 0,03133297871
-4,5 0,07466398507
-5,6 -0,01683176542
-6,7 -0,0090632583
-7,8 0,0030210861
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Splines_biortho/splines [11/15/2003 9:47:17 PM]

Synthesis of Biorthogonal Wavelets
Synthesis of Compactly
Supported
Biorthogonal Wavelets
Synthesis of perfect reconstruction filter
banks
The construction of perfect reconstruction filter banks is simpler
than the construction of conjugate mirror filters because the
quadrature condition is replaced by a Bezout identity:
In particular, spectral factorization is no longer required.
Biorthogonal wavelet synthesis

A theorem by Cohen, Daubechies and Fauveau gives sufficient
conditions for building biorthogonal wavelets.
One quite interesting example is given by biorthogonal spline

wavelets. It is iteresting because it has symmetric scaling functions,
and because there existe a closed form formula for the filters.
The spline example
The h filter is taken to be
with ε=0 if p is even and ε =1 if p is odd. The scaling function is a
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/Synthesis_of_Biortho.html (1 of 3) [11/15/2003 9:47:21 PM]

B-spline of degree p-1 (this can verified by using the recursion

which relates B-splines of different degrees). It is a symmetric
function with respect to 0 if p is odd, and symmetric with respect to
1/2 if p is odd. The corresponding wavelet is respectively
symmetric or antisymmetric. The dual wavelet has p vanishing
moments.
The only constraint on the number of vanishing moments of the

primal wavelet is that it should have the same parity as p. Hence
the symmetries are the same as in the previous case. For
q=(p+p2)/2, the biorthogonal filter h2 of minimum length is given
by
Here is an example for p=3 and p2=7
Biorthogonal cubic B-spline scaling function Dual scaling function

The same filters are used to implement the

dyadic wavelet transform

Construction of Wavelet Bases
Properties Properties
Dilated wavelets are related Properties of biorthogonal
by a scaling equation. wavelets are to be compared
to the orthogonal case.
Rescaling can be interpreted
as discrete filtering.
Scaling equation
Vanishing moments,
support, regularity and As in the orthogonal case, ψ(t) and
symmetry of the wavelet ϕ(t/2) are related by a scaling
equation which is a consequence of
and scaling function are the inclusions of the resolution spaces
determined by the scaling from coarse to fine:
filter.
Scaling equation Similar equations exist for the dual

functions which determine the filters
ψ(t/2) and φ(t) are related by a h2 and g2.
scaling equation, similar to the
equation which relates ϕ(t/2) and Vanishing moments
ϕ(t). It is a consequence of (7.60):
A biorthogonal wavelet has m
vanishing moments if and only if its
dual scaling function generates
polynomials up to degree m. This can
In the Fourier domain, this equation be verified by looking at the
becomes biorthogonal decomposition
formulas.
Hence there is an equivalence
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/Construction_of_wave_bas.html (1 of 5) [11/15/2003 9:47:32 PM]

theorem between vanishing moments

The h and g filters are a conjugate and the number of zeroes of the
mirror filter bank. filter's transfer, provided that duality
has to be taken into account. Thus the
Vanishing moments following three properties are
equivalent:
A wavelet has m vanishing moments
if and only if its scaling function can ● the wavelet ψ has p
generate polynomials of degree
smaller than or equal to m. While this
vanishing moments
property is used to describe the ● the dual scaling
approximating power of scaling function ϕ2 generates
functions, in the wavelet case it has a
"dual" usage, e.g. the possibility to polynomials up to
caracterize the order of isolated degree p
singularities.
● the transfer function of
The number of vanishing moments is the dual filter h2 and it
entirely determined by the p-1 first derivatives
coefficients h[n] of the filter h which
is featured in the scaling equation. vanish at ω=π
If the Fourier transform of the and the dual result is also valid.
wavelet is p continuously Duality appears naturally, because the
differentiable, then the three filters determine the degree of the
following conditions are equivalent: polynomials which can be generated
by the scaling function, and this
● the wavelet ψ has p degree is equal to the number of
vanishing moments of the dual
vanishing moments wavelet.
● the scaling function ϕ
can generate Compact support
polynomials of degree If the filters h et h2 have a finite
smaller than or equal to support, then the scaling functions
p have the same support, and the
● the transfer function of wavelets are compactly supported. If

the supports of the scaling functions
the filter h and its p-1 are respectively [N1,N2] and
first derivatives vanish [M1,M2], then the corresponding
at ω=π. wavelets have support [(N1-
M2+1)/2,(N2-M1+1)/2] and [(M1-
Compact support N2+1)/2,(M2-N1+1)] respectively.
Compactly supported wavelets and

The atoms are thus
scaling functions exist.
compactly supported if and
The scaling function is compactly only if the filters h et h2 are.
supported if and only if the filter h
has a finite support, and their
supports are the same. If the support Regularity
of the scaling function is [N1,N2],
Tchamitchian's theorem provides
then the wavelet support is [(N1-
again a sufficient regularity
N2+1)/2,(N2-N1+1)2]. condition. Remember that this
condition bears on the filter h which
Atoms are thus compactly determines the scaling equation.
supported if and only if the Hence the regularity of the primal
atoms are related to the primal filters.
filter h is.
Daubechies has proved that, to Wavelet balancing
generate an orthogonal wavelet with
Consider the following
p vanishing moment, a filter h with
decomposition of f:
minimum length 2p had to be used.
Daubechies filters, which generate
Daubechies wavelets, have a length
of 2p. The Daubechies filter
coefficients are available as ASCII
The number of vanishing moments of
text files which can be used in a
a wavelet is determined by its dual
spreadsheet, for instance. filter. It corresponds to the
approximating power of the dual
Regularity multiresolution sequence. This is why

it is preferred to synthesize a
Wavelet regularity is much less decomposition filter h with many
important than their vanishing vanishing moments, and possibly
moments. Il is studied in a theorem with a small support.
by Tchamitchian
On the other hand, this same filter h
The following two properties are determines the regularity of ϕ, and
important: hence of ψ. This regularity increases
with the number of vanishing
● there is no compactly supported moments, that is, with the number of
orthogonal wavelet which zeroes of h.
indefinitely differentiable
● for Daubechies wavelets with a Symmetry
large p, the scaling function and
wavelet are l-Lipschitz, where l Unlike the orthogonal case, it is
is of the order of 0.2 p. For large possible to synthesize biorthogonal
classes of orthogonal wavelets, wavelets and scaling functions which
more regularity implies more are symmetric or antisymmetric and
vanishing moments. compactly supported. This makes it
possible to use the folding technique
Meyer wavelets are indefinitely to build wavelets on an interval.
differentiable orthogonal wavelets,
with an infinite support. They are If the filters h and h2 have and odd
generally implemented in the Fourier length and are symmetric with respect
domain. to 0, then the scaling functions have
an even length and are symmetric,
Symmetry and the wavelets are also symmetric.
If the filters have an even length and
Symmetric scaling functions and are symmetric with respect to n=1/2,
wavelets are important because they then the scaling functions are
are used to build bases of regular symmetric with respect to n=1/2,
wavelets over an interval, rather than while the wavelets are antisymmetric.
the real axis. Daubechies has proved
that, for a wavelet to be symmetric or Example
antisymmetric, its filter must have a
linear complex phase, and the only Spline wavelets and scaling functions
symmetric compactly supported are an interesting example of
conjugate mirror filter is the Haar biorthogonal systems. One of the
filter, which corresponds to a scaling functions is a B-spline. A
discontinuous wavelet with one coefficient table is available. There is
vanishing moment. Besides the Haar a general closed form formula for
wavelet, there is no symmetric these filters.
compactly supported orthogonal
wavelet.
Biorthogonal Wavelets and Discrete

Filters
Orthogonal Wavelets and Discrete
Filters

250
Original signal Reconstructed signal
200 using a frame inverse
on dyadic maxima
Reconstructed signal
after a thresholding
150 using a frame inverse
above T=10
on dyadic maxima
100
50
0
0 50 100 150 200 250
Edge Detection
Multiscale Edge Detection

and Reconstruction
As in the one dimensional case, dyadic modulus
maxima are used to dectect edges.
Provided that the two dimensional geometry is
taken into account, these edges can be interpreted
as contours.
A similar algorithm to the one dimensional case
reconstructs a good approximation of an image
from its edges.
Multiscale edges
In images, what is most often perceived as an edge is a curve
across which there is a sharp variation of brightness. To make
things simpler, the image will be assumed to be monochrome.
While the actual concept of an edge is more involved and depends
in particular on a priori knowledge about the featured objects, this
presentation has the advantage of leading to a precise mathematical
definition of an "edge point".
To do so, consider a two dimensioanl wavelet defined by partial

differentiation of a kernel:
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes%20dyadiques/Edge_Detection.html (1 of 4) [11/15/2003 9:47:52 PM]

Edge Detection
The dyadic wavelet transform is defined by
with, for k=1,2,
The two coordinates of the dyadic wavelet transform are that of the
gradient of the convolution of the signal with the dilated kernel:
The multiscale edge points are the points where the dyadic
transform modulus is locally maximum along this direction. This
corresponds to a locally sharpest variation of image intensity
orthogonally to the lines of constant brightness.
Examples
A synthetic example analyzes the edges of a circle.
Another example analyses a classical wavelet picture.
Remark
It is rare that an image line has no hole in it. The brain compensate
these defaults using more elaborate image analysis. Notice that the
use of color is useful. As illustration, here is an optical illusion
where joining edges is far from being obvious:

Edge Detection
Reconstruction
As in the one dimensional case, the frame inverse operator can be
used to reconstruct a minimum norm image with prescribed values
at the maxima locations. Mean square relative errors of l0-2 can be
obtained.
On an example, one can see that the reconstruction error is visually

neglectible.
Implementation
The computations are performed with separable wavelets whose
Fourier transforms are
where g is a finite difference filter; the two wavelets then

approximate the partial derivatives of

Edge Detection
where φ is a scaling function defined by a finite impulse response

filter h. The dyadic wavelet transform is computed by two
dimensional extension of the algorithme à trous.
Back to top
Next Path

Cercle
The original image is on top.

Wavelet
transform Wavelet
Horizontal Vertical Wavelet angle for a transform
wavelet wavelet transform non zero modulus
transform transform modulus modulus maxima
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes%20dyadiques/Circle.html [11/15/2003 9:48:20 PM]

Lena Edges
A high resolution version of the example is available in PDF format (482K).
Original image:
Wavelet
Wavelet transform
transform Wavelet modulus
Horizontal Vertical Wavelet angle for a transform maxima above
wavelet wavelet transform non zero modulus a given
transform transform modulus modulus maxima threshold
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes%20dyadiques/Lena_Edges.html [11/15/2003 9:50:28 PM]

Reconstruction of Lena
Reconstruction from modulus maxima

Original Lena image
and coarse approximation
Reconstruction from thresholded

modulus maxima
and coarse approximation
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes%20dyadiques/Reconstruction_of_Lena.html [11/15/2003 9:51:49 PM]

Wavelets and Filters
Orthogonal Wavelets
and Discrete Filters Biorthogonal Wavelets
g and h are conjugate mirror filters. and Discrete Filters
Conjugate mirror filters are a particular instance of perfect The scaling equations on the scaling functions and
reconstruction filter banks. The dyadic nature of multiresolution wavelets show that the decomposition and reconstruction
approximations are closely related to the possibility of implementing of a signal from a resolution to the next one is
elementary signal subsampling by erasing one sample every two, implemented by perfect reconstruction filter banks.
and elementary oversampling by zero insertion between two
consecutive samples. The scaling equations imply the coefficients a1 [n] and d1
[n] of a signal in Vj and Wj are computed from its
The coefficients 1 [n] and d1 [n] of a signal in Vj and Wj are coefficients a0[n] in Vj-1 by applying the filters h and g
computed from its coefficients a0[n] in Vj-1 by applying conjugate and subsampling the output:
mirror filters and subsampling the output:
a1 [n] = a0 * h1 [2n]
a1 [n] = a0 * h1 [2n]
and
and d1 [n] = a0* g1 [2n] .
d1 [n] = a0* g1 [2n] .
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/wavelets_and_filters.html (1 of 4) [11/15/2003 9:51:54 PM]


In practice this recursion is initialized by considering that
The coefficients of h are defined by the scaling equation
the discrete signal samples are some fine resolution
coefficients.
The coefficients of h and g are defined by the scaling

equations
and the coefficients of g are defined by the wavelet scaling equation

Conversely, a0[n] is reconstructed from a1 [n] and d1 [n]

Conversely, a0[n] is reconstructed from a1 [n] and d1 [n] by inserting by inserting zeroes between two consecutive samples and
summing their convolutions with the dual filters h2 et g2
zeroes between two consecutive samples and summing their
convolutions with h and g: which define the dual scaling equations:
a0 [n] = z(a1) * h [n] + z(d1) * g [n] a0 [n] = z(a1) * h2 [n] + z(d1) * g2 [n]
Wavelets and scaling functions are evaluated as in the orthogonal

case.
The coefficients of h2 and g2 are defined by the scaling

Hence the construction of orthogonal
equations
wavelets is equivalent to the synthesis of
conjugate mirror filters having a stability
property.
We will concentrate on Finite Impulse Response
filters, e.g., to compactly supported wavelets.
Filtering
This algorithm is used to evaluate the scaling functions

and wavelets. Indeed the coefficients of a scaling function
(resp. wavelet) are all zero but for one within their on

resolution (resp. detail) space basis. The reconstruction

algorithm provides the coefficients in the finer
resolutions. For high resolutions, the scaling coefficients
are considered to be samples of the function.
Hence the construction of

biorthogonal wavelets is
equivalent to the synthesis of
perfect reconstruction filters
having a stability property.
Filtering

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/MallatMeyerUS.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/MallatMeyerUS.gif [11/15/2003 9:51:55 PM]

Filter Synthesis
Synthesis of Synthesis of Compactly

Compactly Supported
Supported Biorthogonal Wavelets
Orthogonal Synthesis of perfect reconstruction filter
Wavelets banks
Synthesis of conjugate The construction of perfect reconstruction filter banks is simpler
mirror filters than the construction of conjugate mirror filters because the
quadrature condition is replaced by a Bezout identity:
A finite impulse congugate mirror
filter bank is caracterized by a filter h
which satisfies
In particular, spectral factorization is no longer required.
Biorthogonal wavelet synthesis

where h(ω) is a trigonometric A theorem by Cohen, Daubechies and Fauveau gives sufficient
polynomial (or, equivalently, a conditions for building biorthogonal wavelets.
polynomial with respect to the shift
and sign operators). The π frequency One quite interesting example is given by biorthogonal spline
shift amounts to a change of sign wavelets. It is iteresting because it has symmetric scaling
every two coefficient. Moreover, this functions, and because there existe a closed form formula for the
continuous time transfer must vanish filters.
up to the order p at π in order to have
a wavelet with p vanishing moments. The spline example
The synthesis of such filters can be The h filter is taken to be

done using several methods. The best
known filters are Daubechies's
compactly supported filters. An
outline of the construction method is
available. with ε=0 if p is even and ε =1 if p is odd. The scaling function is a
B-spline of degree p-1 (this can verified by using the recursion
Wavelet synthesis which relates B-splines of different degrees). It is a symmetric
function with respect to 0 if p is odd, and symmetric with respect
There remains to check that the filter to 1/2 if p is odd. The corresponding wavelet is respectively
h does generate a scaling function. To symmetric or antisymmetric. The dual wavelet has p vanishing
do so, it is enough to verify that the moments.
transfer of h does not vanish on [-
π/2,π/2] (theorem by Mallat and The only constraint on the number of vanishing moments of the
Meyer). The construction on primal wavelet is that it should have the same parity as p. Hence
Daubechies'compactly supported the symmetries are the same as in the previous case. For
orthogonal wavelets is presented q=(p+p2)/2, the biorthogonal filter h2 of minimum length is given
here. by
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/Synthesis_of_Filters.html (1 of 2) [11/15/2003 9:52:02 PM]

Filter Synthesis
The coefficients of the

conjugatemirror filter h can be
obtained in Wavelab, are freeware
Matlab toolbox, using the function
MakeONFilter. Here is an example for p=3 and p2=7
The same filters are used to

implement the
dyadid wavelet transform
Biorthogonal cubic B-spline scaling function Dual scaling function

dyadic wavelet transform
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/Synthesis_of_Filters.html (2 of 2) [11/15/2003 9:52:02 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/AMRDefUS.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/AMRDefUS.gif [11/15/2003 9:52:06 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Ondelettes/WTHeis.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Ondelettes/WTHeis.gif [11/15/2003 9:52:07 PM]

Multiresolution Examples
Examples
The simplest multiresolution example is the Haar multiresolution.
In this case, θ is the characteristic function of [0,1], and the basis
coefficients in a resolution space are computed by an average on a
suitable interval. You can see an example of a Haar multiresolution
of classical painting (Pan et Syrinx by Boucher, 1759, National
Gallery, London).
The sequence of polynomial splines spaces with steps 2j , j in Z, is

a multiresolution approximation.
A cubic B-spline and its Fourier transform.
The following movie (in QuickTime format) shows the succesive

approximations of a wavelet classics.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Multiresolution_examples.html (1 of 2) [11/15/2003 9:52:09 PM]

Multiresolution Examples
If you browse cannot display QuickTime movies, an image per image display is
available.
Back to Multiresolution Approximations
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Multiresolution_examples.html (2 of 2) [11/15/2003 9:52:09 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/BoucherHaar.gif
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/BoucherHaar.gif [11/15/2003 9:53:39 PM]

Lena
see resolution:
1234
See all resolutions together
Back to examples
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/LenaUS.html [11/15/2003 9:53:47 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Lena1.GIF
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Lena1.GIF [11/15/2003 9:53:49 PM]



Quatre Lenas
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Quatre_Lenas.html [11/15/2003 9:54:58 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/?C=N&O=D [11/15/2003 9:55:00 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/?C=M&O=A [11/15/2003 9:55:03 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/?C=S&O=A [11/15/2003 9:55:05 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/?C=D&O=A [11/15/2003 9:55:09 PM]

Index of /~chaplais/Wavetour_presentation/Tables
Index of
/~chaplais/Wavetour_presentation/Tables
Parent Directory -
Daubechies/ 08-Jul-1998 01:42 -
Daubechies_coefficie..> 01-Jul-1998 15:29 2.6K
Splines_biortho/ 08-Jul-1998 01:42 -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/ [11/15/2003 9:55:13 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/Daubechies_coefficients
n h_p[n]
p = 2 0 ,482962913145
1 ,836516303738
2 ,224143868042
3 -,129409522551
p = 3 0 ,332670552950
1 ,806891509311
2 ,459877502118
3 -,135011020010
4 -,085441273882
5 ,035226291882
p = 4 0 ,230377813309
1 ,714846570553
2 ,630880767930
3 -,027983769417
4 -,187034811719
5 ,030841381836
6 ,032883011667
7 -,010597401785
p = 5 0 ,160102397974
1 ,603829269797
2 ,724308528438
3 ,138428145901
4 -,242294887066
5 -,032244869585
6 ,077571493840
7 -,006241490213
8 -,012580751999
9 ,003335725285
p = 6 0 ,111540743350
1 ,494623890398
2 ,751133908021
3 ,315250351709
4 -,226264693965
5 -,129766867567
6 ,097501605587
7 ,027522865530
8 -,031582039317
9 ,000553842201
10 ,004777257511
11 -,001077301085
p = 7 0 ,077852054085
1 ,396539319482
2 ,729132090846
3 ,469782287405
4 -,143906003929
5 -,224036184994
6 ,071309219267
7 ,080612609151
8 -,038029936935
9 -,016574541631
10 ,012550998556
11 ,000429577973
12 -,001801640704
13 ,000353713800
p = 8 0 ,054415842243
1 ,312871590914
2 ,675630736297
3 ,585354683654
4 -,015829105256
5 -,284015542962
6 ,000472484574
7 ,128747426620
8 -,017369301002
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/Daubechies_coefficients (1 of 2) [11/15/2003 9:55:27 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/Daubechies_coefficients
9 -,044088253931
10 ,013981027917
11 ,008746094047
12 -,004870352993
13 -,000391740373
14 ,000675449406
15 -,000117476784
p = 9 0 ,038077947364
1 ,243834674613
2 ,604823123690
3 ,657288078051
4 ,133197385825
5 -,293273783279
6 -,096840783223
7 ,148540749338
8 ,030725681479
9 -,067632829061
10 ,000250947115
11 ,022361662124
12 -,004723204758
13 -,004281503682
14 ,001847646883
15 ,000230385764
16 -,000251963189
17 ,000039347320
p = 10 0 ,026670057901
1 ,188176800078
2 ,527201188932
3 ,688459039454
4 ,281172343661
5 -,249846424327
6 -,195946274377
7 ,127369340336
8 ,093057364604
9 -,071394147166
10 -,029457536822
11 ,033212674059
12 ,003606553567
13 -,010733175483
14 ,001395351747
15 ,001992405295
16 -,000685856695
17 -,000116466855
18 ,000093588670
19 -,000013264203
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/Daubechies_coefficients (2 of 2) [11/15/2003 9:55:27 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/Daubechies_coefficients.txt
n h_p[n]
p = 2 0 ,482962913145
1 ,836516303738
2 ,224143868042
3 -,129409522551
p = 3 0 ,332670552950
1 ,806891509311
2 ,459877502118
3 -,135011020010
4 -,085441273882
5 ,035226291882
p = 4 0 ,230377813309
1 ,714846570553
2 ,630880767930
3 -,027983769417
4 -,187034811719
5 ,030841381836
6 ,032883011667
7 -,010597401785
p = 5 0 ,160102397974
1 ,603829269797
2 ,724308528438
3 ,138428145901
4 -,242294887066
5 -,032244869585
6 ,077571493840
7 -,006241490213
8 -,012580751999
9 ,003335725285
p = 6 0 ,111540743350
1 ,494623890398
2 ,751133908021
3 ,315250351709
4 -,226264693965
5 -,129766867567
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/Daubechies_coefficients.txt (1 of 4) [11/15/2003 9:55:28 PM]

6 ,097501605587
7 ,027522865530
8 -,031582039317
9 ,000553842201
10 ,004777257511
11 -,001077301085
p = 7 0 ,077852054085
1 ,396539319482
2 ,729132090846
3 ,469782287405
4 -,143906003929
5 -,224036184994
6 ,071309219267
7 ,080612609151
8 -,038029936935
9 -,016574541631
10 ,012550998556
11 ,000429577973
12 -,001801640704
13 ,000353713800
p = 8 0 ,054415842243
1 ,312871590914
2 ,675630736297
3 ,585354683654
4 -,015829105256
5 -,284015542962
6 ,000472484574
7 ,128747426620
8 -,017369301002
9 -,044088253931
10 ,013981027917
11 ,008746094047
12 -,004870352993
13 -,000391740373

14 ,000675449406
15 -,000117476784
p = 9 0 ,038077947364
1 ,243834674613
2 ,604823123690
3 ,657288078051
4 ,133197385825
5 -,293273783279
6 -,096840783223
7 ,148540749338
8 ,030725681479
9 -,067632829061
10 ,000250947115
11 ,022361662124
12 -,004723204758
13 -,004281503682
14 ,001847646883
15 ,000230385764
16 -,000251963189
17 ,000039347320
p = 10 0 ,026670057901
1 ,188176800078
2 ,527201188932
3 ,688459039454
4 ,281172343661
5 -,249846424327
6 -,195946274377
7 ,127369340336
8 ,093057364604
9 -,071394147166
10 -,029457536822
11 ,033212674059
12 ,003606553567
13 -,010733175483
14 ,001395351747

15 ,001992405295
16 -,000685856695
17 -,000116466855
18 ,000093588670
19 -,000013264203

Synthesis of Compactly Supported Orthogonal Wavelets
Synthesis of Compactly
Supported
Orthogonal Wavelets
Synthesis of conjugate mirror filters
A finite impulse congugate mirror filter bank is caracterized by a
filter h which satisfies
where h(ω) is a trigonometric polynomial (or, equivalently, a

polynomial with respect to the shift and sign operators). The π
frequency shift amounts to a change of sign every two coefficient.
Moreover, this continuous time transfer must vanish up to the order
p at π in order to have a wavelet with p vanishing moments.
The synthesis of such filters can be done using several methods.

The best known filters are Daubechies's compactly supported
filters. An outline of the construction method is available.
Wavelet synthesis
There remains to check that the filter h does generate a scaling
function. To do so, it is enough to verify that the transfer of h does
not vanish on [-π/2,π/2] (theorem by Mallat and Meyer). The
construction on Daubechies'compactly supported orthogonal
wavelets is presented here.
The coefficients of the conjugatemirror filter h can be obtained in

Wavelab, are freeware Matlab toolbox, using the function
MakeONFilter.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/Synthesis_of_Ortho.html (1 of 2) [11/15/2003 9:55:36 PM]

Synthesis of Compactly Supported Orthogonal Wavelets

dyadid wavelet transform
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/Synthesis_of_Ortho.html (2 of 2) [11/15/2003 9:55:36 PM]

Biorthogonal Multiresolution Approximations
Biorthogonal
A pair [(Vj),(V*l)] of multiresolution approximations is a
biorthogonal multiresolution system if and only if
Then V*0 has a Riesz basis of the form θ∗(t-n), n in Z, such that
the translations of θ and of θ∗ form a biorthogonal system:
We have a biorthogonal bases system instead of a single orthogonal

basis.
Example
Below is a pair of biorthogonal scaling functions. The first one is a

cubic B-spline. Since there is no orthogonality constraint, the B-
spline can be kept as an atom et preserve thus both symmetry and
compact support.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Biorthogonal_Multires.html (1 of 2) [11/15/2003 9:55:41 PM]

Biorthogonal Multiresolution Approximations
Back to the Properties of Multiresolution Approximations
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Biorthogonal_Multires.html (2 of 2) [11/15/2003 9:55:41 PM]

Strang and Fix Paradox
Yes, because the intersection of the resolution spaces is taken in

L2; but polynomials do not have a finite energy. Hence, trying to
analyse a polynomial of degree n with (orthogonal, to symplify)
wavelets with p>=n vanishing moments:
yields a zero result. On the other hand, the decomposition
does reconstruct the polynomial, with zero coefficient on the

second line of the decomposition.
In fact, the second formula is more general and can represent

functions which do not belong to L2.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Multiresolution/Strang_and_Fix_Paradox.html [11/15/2003 9:55:42 PM]

Daubechies Filters (part 1)
This presentation is inspired from the Daubechies classic, "Ten

Lectures on Wavelets", SIAM, 1992.
Daubechies gives the following result:
Hence can easily be found. There remains to find h. Observe

that is symmetric with respect to the pulsation. Daubechies uses
then the spectral factorization theorem:
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/Daubechies_Filters.html (1 of 2) [11/15/2003 9:55:46 PM]

Daubechies Filters (part 1)
which leads to h.
The "classical" Daubechies filters are such that R=0; in the spectral
factorization, only the zeroes within the unit circle are kept.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/Daubechies_Filters.html (2 of 2) [11/15/2003 9:55:46 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/CDFUS.GIF
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/CDFUS.GIF [11/15/2003 9:55:49 PM]

MakeONFilter
MakeONFilter -- Generate
Orthonormal QMF Filter for
Wavelet Transform
Usage
qmf = MakeONFilter(Type,Par)
Inputs
Type
string, 'Haar', 'Beylkin', 'Coiflet',

'Daubechies', 'Symmlet', 'Vaidyanathan'
Par
integer, e.g. if Type = 'Coiflet', Par=3

specifies a Coiflet-3 wavelet
Outputs
qmf quadrature mirror filter
Description
The Haar filter (which could be considered a
Daubechies-2) was the first wavelet, though not
called as such, and is discontinuous.
The Beylkin filter places roots for the

frequency response function close to the Nyquist
frequency on the real axis.
The Coiflet filters are designed to give both

the mother and father wavelets 2*Par vanishing
moments; here Par may be one of 1,2,3,4 or 5.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/MakeONFilter.html (1 of 2) [11/15/2003 9:55:54 PM]

MakeONFilter
The Daubechies filters maximize the smoothness

of the father wavelet (or "scaling function") by
maximizing the rate of decay of its Fourier
transform. They are indexed by their length,
Par, which may be one of 4,6,8,10,12,14,16,18 or
20.
Symmlets are the "least asymmetric" compactly-

supported wavelets with maximum number of
vanishing moments, here indexed by Par, which
ranges from 4 to 10.
The Vaidyanathan filter gives an exact

reconstruction, but does not satisfy any moment
condition. The filter has been optimized for
speech coding.
See Also
FWT_PO, IWT_PO, FWT2_PO, IWT2_PO, WPAnalysis
References
The books by Daubechies and Wickerhauser.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/filtres/MakeONFilter.html (2 of 2) [11/15/2003 9:55:54 PM]

Examples of Fourier Ridges
Examples of Wavelet Ridges

Below are the wavelet ridges of the sum of parallel linear chirps.
The frequency tracking gets worse as the frequencies increase. This

is because the frequency resolution of the wavelet transform
decreases when the frequency increases. Indeed, for these linear
chirps, the relative frequency difference goes to 0 as t increases,
and this creates "interferences" between the ridges. Similarly,
interferences can be observed in the figure below, at t=900,
between the linear chirp and the lower frequency Gabor chirp.
On the contrary, the increase of time resolution as the frequency

increases makes it possible to track the instantaneous frequencies
of the hyperbolic chirps:
http://cas.ensmp.fr/~chaplais/Wavetour_present...ees/Ondelettes/Examples_of_Wavelet_Ridges.html (1 of 2) [11/15/2003 9:55:56 PM]

Examples of Fourier Ridges
Conclusion on Instantaneous Frequency Detection
http://cas.ensmp.fr/~chaplais/Wavetour_present...ees/Ondelettes/Examples_of_Wavelet_Ridges.html (2 of 2) [11/15/2003 9:55:56 PM]

Conclusion on the Detection of Instantaneous Frequencies
Conclusion on the Detection

of Instantaneous Frequencies
The illusion of intantaneous frequencies
Because of the Heisenberg uncertainty theorem, there is no
intrinsic definition of the instantaneous frequencies of a finite
energy signal.
The structure of the time frequency tiling achieved by the atoms of

a time frequency transform determines its time frequency
resolution.
To detect the instantaneous frequencies of a signal, an adapted

time frequency transform must be used.
Towards more general tilings

Adapting the transforms is one of the reasons why time frequency
localized transforms have been generalized to more general time
frequency tilings.
Wavelet packets and local cosine transforms are such

generalizations. They are explained in chapter 8 of the book, but
are not described here.
Optimal basis search

These transforms correspond to families of bases which can be
represented as a maximal tree of time frequency refinements.
This structure is suited to the search for an "optimal" orthogonal

basis using a dynamic programming algorithm within the tree.
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Conclusion_on_detection.html (1 of 2) [11/15/2003 9:55:58 PM]

Conclusion on the Detection of Instantaneous Frequencies
If non orthogonal bases are allowed, the additivity of the cost is

lost. Basis and matching pursuits are example of methods for a
"best" basis selection.
These best basis selections are explained in chapter 9 of the book.
While this leads theoretically to a compact representation of the

signal, the cost of the basis coding has to be taken into account.
Moreover, too much relaxation on the basis requirements may lead
to a representation whithout any structure information, because the
basis is too much taylored on the signal.
Back to the top
Next topic
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Conclusion_on_detection.html (2 of 2) [11/15/2003 9:55:58 PM]

Original
image
Wavelet
transform
Wavelet Wavelet Wavelet Wavelet Wavelet modulus
transform transform transform transform transform maxima
along the along the modulus angle for a modulus after some
horizontal vertical non zero maxima thresholding
direction direction modulus
Examples of Windowed Fourier

Ridges
Below are the windowed Fourier ridges of the sum of parallel
linear chirps.
The two frequencies are well detected.
Now the ridges of a sum of chirps:
Between t=500 and t=600, the transform fails to discriminate the

frequencies because of its limited frequency resolution.
The limited time resolution at high frequencies explains why the

windowed Fourier transforms fails to track the instantaneous
frequencies of the sum of two hyperbolic chirps:
http://cas.ensmp.fr/~chaplais/Wavetour_presenta...Fourier_Fenetre/Examples_of_Fourier_Ridges.html (1 of 2) [11/15/2003 9:59:20 PM]

Indeed, the instantaneous frequencies vary too much within the

Fourier window when close to the asymptote.
Conclusion on Instantaneous Frequency Detection
http://cas.ensmp.fr/~chaplais/Wavetour_presenta...Fourier_Fenetre/Examples_of_Fourier_Ridges.html (2 of 2) [11/15/2003 9:59:20 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies/?C=S&O=D [11/15/2003 9:59:24 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/?C=N&O=D [11/15/2003 9:59:29 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/?C=M&O=A [11/15/2003 9:59:31 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/?C=S&O=A [11/15/2003 9:59:32 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/?C=D&O=A [11/15/2003 9:59:35 PM]

A Wavelet Tour of Signal Processing (a Presentation)
A WAVELET TOUR
BY STÉPHANE MALLAT
A SHORT PRESENTATION BY F.
CHAPLAIS
Version française
US-English version
( Full Strang and Fix conditions added, with proof!)
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ [11/15/2003 9:59:44 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies_coefficients.txt
n h_p[n]
p = 2 0 ,482962913145
1 ,836516303738
2 ,224143868042
3 -,129409522551
p = 3 0 ,332670552950
1 ,806891509311
2 ,459877502118
3 -,135011020010
4 -,085441273882
5 ,035226291882
p = 4 0 ,230377813309
1 ,714846570553
2 ,630880767930
3 -,027983769417
4 -,187034811719
5 ,030841381836
6 ,032883011667
7 -,010597401785
p = 5 0 ,160102397974
1 ,603829269797
2 ,724308528438
3 ,138428145901
4 -,242294887066
5 -,032244869585
6 ,077571493840
7 -,006241490213
8 -,012580751999
9 ,003335725285
p = 6 0 ,111540743350
1 ,494623890398
2 ,751133908021
3 ,315250351709
4 -,226264693965
5 -,129766867567
6 ,097501605587
7 ,027522865530
8 -,031582039317
9 ,000553842201
10 ,004777257511
11 -,001077301085
p = 7 0 ,077852054085
1 ,396539319482
2 ,729132090846
3 ,469782287405
4 -,143906003929
5 -,224036184994
6 ,071309219267
7 ,080612609151
8 -,038029936935
9 -,016574541631
10 ,012550998556
11 ,000429577973
12 -,001801640704
13 ,000353713800
p = 8 0 ,054415842243
1 ,312871590914
2 ,675630736297
3 ,585354683654
4 -,015829105256
5 -,284015542962
6 ,000472484574
7 ,128747426620
8 -,017369301002
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies_coefficients.txt (1 of 2) [11/15/2003 9:59:46 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies_coefficients.txt
9 -,044088253931
10 ,013981027917
11 ,008746094047
12 -,004870352993
13 -,000391740373
14 ,000675449406
15 -,000117476784
p = 9 0 ,038077947364
1 ,243834674613
2 ,604823123690
3 ,657288078051
4 ,133197385825
5 -,293273783279
6 -,096840783223
7 ,148540749338
8 ,030725681479
9 -,067632829061
10 ,000250947115
11 ,022361662124
12 -,004723204758
13 -,004281503682
14 ,001847646883
15 ,000230385764
16 -,000251963189
17 ,000039347320
p = 10 0 ,026670057901
1 ,188176800078
2 ,527201188932
3 ,688459039454
4 ,281172343661
5 -,249846424327
6 -,195946274377
7 ,127369340336
8 ,093057364604
9 -,071394147166
10 -,029457536822
11 ,033212674059
12 ,003606553567
13 -,010733175483
14 ,001395351747
15 ,001992405295
16 -,000685856695
17 -,000116466855
18 ,000093588670
19 -,000013264203
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Daubechies_coefficients.txt (2 of 2) [11/15/2003 9:59:46 PM]

Index of /~chaplais/Wavetour_presentation/Tables/Splines_biortho
Index of
/~chaplais/Wavetour_presentation/Tables/Splines_biortho
Parent Directory -
splines 04-Jun-1999 12:56 653
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Splines_biortho/ [11/15/2003 9:59:50 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/?C=M&O=D [11/15/2003 9:59:53 PM]

Index of
Parent Directory -
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/?C=S&O=D [11/15/2003 9:59:54 PM]

Une presentation du livre de Stephane Mallat
A WAVELET TOUR
PAR STÉPHANE MALLAT
Disponible en français!
PRÉSENTÉ SUCCINCTEMENT PAR F.

CHAPLAIS
ce lien est pour ceux qui n'aiment pas les préambules
Configuration souhaitée
Ce site a été testé sous Netscape Navigator 2 et 3. La version 2 ne restitue pas certains
symboles mathématiques en police Symbol, mais en une quantité négligeable pour la
compréhension de l'exposé. Un écran en 256 niveau de gris est nécessaire (traitement
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Wavetour_presentation_fr.html (1 of 4) [11/15/2003 9:59:59 PM]

d'image oblige), et certain graphiques sont en couleurs pour mieux comparer les
courbes. La plupart des images sont au format GIF entrelacé avec transparence sur le
blanc.
Avertissement
Cette présentation s'inspire du livre de S. Mallat, elle ne prétend pas en être un reflet
exact. Elle s'intéresse plus particulièrement aux sujets suivants:
● Analyse de Fourier (chapitre 2)

● Analyse temps-fréquence (chapitre 4, excepté les énergies
quadratiques temps-fréquence))
● Les frames (chapitre 5)
● Analyse de singularités et reconstruction (chapitre 6, excepté les
multifractales)
● Bases d'ondelettes et bancs de filtres (chapitre 7)
Sont omis:
● le chapitre 3 sur l'échantillonnage (sauf les algorithmes de FFT et

de convolutions rapides, qui sont brièvement décrits)
● le chapitre 8 sur les paquets d'ondelettes et les bases de cosinus
locaux
● le chapitre 9 sur l'approximation
● le chapitre 10 sur l'estimation (qui est en train d'être réécrit par S.
Mallat)
● le chapitre 11 sur le codage (qui sera probablement rajouté si j'en
ai le temps)
Une table des matières provisoire de la version française est disponible.
Chemins de parcours
Quatre parcours sont proposés, couvrant quatre types de sujets. Ces

sujets s'enchaînent. De nombreuses passerelles existent par ailleurs
de l'un à l'autre.
● analyse fréquentielle
● des ondelettes dyadiques aux filtres
● des filtres aux ondelettes dyadiques

● analyse de régularité
● frames
Index
Pour un accès direct, voici une liste de liens menant aux principaux sujets:
Transformée de Fourier
Fréquence instantanée d'un signal analytique
Localisation temps-fréquence
Transformée de Fourier fenêtrée et transformée en ondelettes
Frames et bases de Riesz
Frames de Fourier fenêtrées et d'ondelettes
Approximations multirésolutions
Bases d'ondelettes
Bancs de filtres
Analyse de la régularité d'un signal
Détection de singularités
Reconstruction à partir des maxima
Détection de contours dans l'image et reconstruction
Algorithmes:
Transformée de Fourier rapide et convolutions

Transformée de Fourier fenêtrée rapide
Transformée en ondelettes rapide
Transformée dyadique rapide
Décomposition et reconstruction sur des bases
d'ondelettes orthonormées
Calculs numériques
Toutes les illustrations numériques du livre ont été réalisées sous Wavelab, une
Toolbox freeware pour Matlab, disponible à l'adresse

http://www-stat.stanford.edu/~wavelab/
Uvi Wave est une autre toolbox Matlab en freeware.
Des travaux pratiques sont également disponibles. Ils ne nécessitent qu'un logiciel
compatible Matlab 4 (aucune toolbox n'est nécessaire) et ne requièrent aucune
connaissance préalable. Tout en se familiarisant avec Matlab, on implémente des
filtres miroirs conjugués et et on les utilise pour comparer diverses méthodes
d'approximation.
Historique:
4 Juin 1999: ajout des conditions générales de Stang et Fix, avec la preuve!
18 Mai 2001: ajout de la transition des bancs de filtres vers les analyses
multirésolution.
Dernière mise à jour: 7 Mars 2001
Merci de me faire part de vos remarques.
Centre
Automatique
et Systèmes
Recherche
options

Index of
Parent Directory -
splines 04-Jun-1999 12:56 653
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Splines_biortho/?C=N&O=D [11/15/2003 10:00:00 PM]

Index of
Parent Directory -
splines 04-Jun-1999 12:56 653
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Splines_biortho/?C=M&O=A [11/15/2003 10:00:02 PM]

Index of
Parent Directory -
splines 04-Jun-1999 12:56 653
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Splines_biortho/?C=S&O=A [11/15/2003 10:00:03 PM]

Index of
Parent Directory -
splines 04-Jun-1999 12:56 653
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Tables/Splines_biortho/?C=D&O=A [11/15/2003 10:00:04 PM]

Decomposition and Reconstruction Schemes
Decomposition
Reconstruction
http://cas.ensmp.fr/~chaplais/Wavetour_presentation/ondelettes/Decomposition_and_Reconst.html [11/15/2003 10:00:05 PM]

Wavelettour

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Wavelettour

Uploaded by

Copyright:

Available Formats

A wavelet tutorial from S.

for those who hate preambles

Hardware and Software Requirements

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Wavetour_presentation_US.html (1 of 4) [11/15/2003 9:42:50 PM]

● Fourier analysis (chapter 2)

● chapter 3 about discrete signals (except for the FFT and

Four tours are proposed, corresponding to four different topics.

Fast Fourier Transforms and Convolutions

Uvi Wave is another freeware Matlab Toolbox.

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Wavetour_presentation_US.html (3 of 4) [11/15/2003 9:42:50 PM]

Last update: February 2, 1999

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Wavetour_presentation_US.html (4 of 4) [11/15/2003 9:42:50 PM]

Period and Frequency

A frequency is the inverse of a period.

A one dimensional signal is T>0 periodic if it is unchanged by a

In the representation over an interval of length T, the

As a first approximation, a music piece may be modelized as a

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Intro/Periodicity_US.html (1 of 2) [11/15/2003 9:42:52 PM]

determined by the tempo. Every elementary signal can be

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Intro/Periodicity_US.html (2 of 2) [11/15/2003 9:42:52 PM]

This analysis seems to fits the needs of music coding by using a

● there may be points where the Fourier series does not

A solution consists in using a transform where the frequency is

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Fourier_Series.html [11/15/2003 9:42:54 PM]

The inverse Fourier transform represents f as a sum of sinusoids

This makes it possible for the Fourier transform to give indications

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Fourier_Transform.html (1 of 2) [11/15/2003 9:42:57 PM]

on the regularity of a signal.

There exist, however, a definition of the instantaneous frequency of

It is useless in practice because it fails to detect the summation of

Trying to discriminate each of the stacked frequencies leads to a

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Fourier_Transform.html (2 of 2) [11/15/2003 9:42:57 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Agebraic_Properties.html [11/15/2003 9:43:00 PM]

This property is extended to Lipschitz regularity:

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/Regularity_with_Fourier.html [11/15/2003 9:43:13 PM]

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Regularite/LipschitzDefUS.GIF [11/15/2003 9:43:15 PM]

Fast Fourier Transform

Using this formula to compute the Fourier transform requires N2

A simple computation shows that the even frequency coefficients

fp[n] = f[n] + f[n+N/2]

fi[n] = (f[n]-f[n+N/2]) e-2iπn/N

One verifies by induction that the number of operations required by

This is the basic principle of the Fast Fourier Transform. Variants

Convolutions and circular convolutions

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/FFTUS.html (1 of 2) [11/15/2003 9:43:21 PM]

The discrete Fourier transform of a circular convolution is the

For two signals f and h with a length M>=32, computing their

for 0<=n<2M. The circular convolution is itself computed by

Fast Windowed Fourier Transform

Fast Wavelet Transform

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/transformees/Fourier/FFTUS.html (2 of 2) [11/15/2003 9:43:21 PM]

and fa can properly be decomposed into a module and a complex

The instantanneous frequency is the nonnegative derivative of the

For a sinusoidal wave, this definition coincides with the usual

Hence the caracterization of the instaneous frequency of a signal in

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Intro/Instantanous_Frequency.html (1 of 2) [11/15/2003 9:43:29 PM]

requires other mathematical tools.

However, analytic signals are very useful in the synthesis of signals

http://cas.ensmp.fr/~chaplais/Wavetour_presentation/Intro/Instantanous_Frequency.html (2 of 2) [11/15/2003 9:43:29 PM]

The Fourier transform can be viewed as a representation of a

To represent the frequency behavior of a signal locally in time, the

This time-frequency localization is limited by the following two