See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/269743218

Applications and Properties of Different Kind of Wavelet

Article in Indian Journal of Industrial and Applied Mathematics · January 2013

DOI: 10.5958/j.1945-919X.4.1a.001

CITATIONS READS
2 927

2 authors, including:

Ratikanta Behera
Indian Institute of Science
47 PUBLICATIONS 560 CITATIONS

SEE PROFILE

All content following this page was uploaded by Ratikanta Behera on 25 October 2017.

The user has requested enhancement of the downloaded file.

 INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS 1
Vol. 4, Issue 1, Jan-June 2013, pp- 1-15
DOI: 10.5958/j.1945-919X.4.1a.001

Applications and Properties of Different Kind of

Wavelet

Mani Mehra* and Ratikanta Behera

Department of Mathematics, Indian Institute of Technology, New Delhi - 110016, India

Abstract: In this paper, we present an overview of different kinds of wavelet and its properties. We also
point out advantages and disadvantages of differents kinds of wavelet.
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

Keywords: Wavelets, Multiresolution analysis, Vanishing Moment

Members Copy, Not for Commercial Sale
www.IndianJournals.com

1 Introduction of Wavelet
The main aim of the discovery of wavelet is study the time-frequency signal analysis.
The Fourier transform analysis can’t analyze local information of a signals, means it does
not reflect the change of wavenumber with space or frequency with time and it also can’t
analyze physical situation of a signal where the signal contains discontinuities and sharp
spikes. This idea of wavelet is not new, because early of 1800’s Joseph Fourier discovered
the idea of the approximation of a function using sines and cosines functions. However,
in wavelet analysis, the scale that we use to see a function plays a special role. Wavelet
algorithms process function at different scales or resolutions. The beauty of wavelet is if
we look at a signal or function with a “large window”, we would predict large features.
Similarly, if we look at a signal with a “small window”, we would predict small features.
This construction of wavelets is very interesting and useful. For many decades, Engi-
neers and Scientists wanted to approximate the functions more accurately, but in Fourier
analysis, signals or function is approximate by using the bases of sines and cosine. These
functions are non-local (and stretch out to infinity). Therefore this is a very poor job in
approximating sharp spikes. But with wavelet analysis, we can approximating a function
or a signal with sharp discontinuities. In other way wavelets are a set of non-linear bases.
When projecting (or approximating) a function in terms of wavelets, the wavelet basis
functions are chosen according to the function being approximated.
∗
E-mail:mmehra@maths.iitd.ac.in
†
Email:ratikanta ma@student.iitd.ac.in

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 2 Mani Mehra and Ratikanta Behera

Wavelets were developed independently in the fields of mathematics, quantum physics,

electrical engineering [1], and seismic geology [2]. There are many more applications
such as image compression [3], astronomy, acoustics, nuclear engineering, neurophysiol-
ogy, music, magnetic resonance imaging, speech discrimination [4], human vision, radar,
turbulence [5, 6], earthquake prediction [7].
In 1982 Jean Morlet and Alex Grossman [8] introduced the idea of wavelets, which is
a family of functions constructed by using translation and dilation of a single function
[9], which is called mother wavelet. Mathematically, wavelets are mathematical functions
that cut up date into different frequency components, and then study each component
with a resolution matched to its scale [9]. Now we will discuss varieties of wavelet.

2 Different kinds of wavelet

2.1 Daubechies wavelet
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

Daubechies wavelets are a family of orthogonal wavelets defining a discrete wavelet

Members Copy, Not for Commercial Sale

transform and characterized by a maximal number of vanishing moments for some given
www.IndianJournals.com

support. With each wavelet type of this class, there is a scaling function which generates
an orthogonal multi-resolution analysis (MRA) [10, 11]. By using MRA , I. Daubechies [9]
show that for any non negative integer n there axist an orthogonal compactly supported
wavelet such that all that derivatives upto order n exist. MRA is characterized by the
following axims:
• V j ⊂ V j+1

∞
• V j is dense in L2 (R),
j=−∞

• f (x) ∈ V j if and only if f (2x) ∈ V j+1 for all j ∈ Z,

• there exist a function φ ∈ V 0 , such that{φ0k = φ(x − k), k ∈ Z}
Now we define W j is orthogonal compliment of V j in V j+1 , i.e. V j ⊥ W j and

V j+1 = V j ⊥ W j (1)

Then there exist a scaling function φ(x) such that φjk (x) = 2j/2 φ(2j x−k) is an orthogonal
basis for V j and similarly there exist ψkj (x) = 2j/2 ψ(2j x − k) is an orthogonal basis for
W j . Since the space V 0 ⊂ V 1 , any function in V 0 can be expanded in terms of the basis
functions of V 1 . In particular, φ(x) = φ00 (x) ∈ V 0 , so

∞
φ(x) = ak φ1k (x) (2)
k=−∞

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 Applications and Properties of different Kind of Wavelet 3

The equation (2) is called as dilation equation (two scale relation for scaling function)
and for Daubechies compactly supported scaling function only finitely many ak , k =

0, 1, ..., D − 1 will be non zero. Where D is the even positive integer called support of
wavelet and a0 , a1 , a2 , ..., aD−1 are called low pass filter coefficients. Similarly, Daubechies
compactly supported wavelet ψ(x) ∈ W 0 ⊂ V 1 , so


D−1
ψ(x) = bk φ1k (x) (3)
k=0

equation (3) is called wavelet equation (two scale relation for wavelet function) and
b0 , b1 , b2 , ..., bD−1 are called low pass filter coefficients, and these filter coefficients are
connected by the relation bk = (−1)k aD−1−k , k = 0, 1, 2, ...D − 1.

db2: φ db2:ψ
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

1.5 2
Members Copy, Not for Commercial Sale

1 1
www.IndianJournals.com

0.5 0

0 −1

−0.5 −2
0 1 2 3 0 1 2 3

Figure 1:

Daubechies orthogonal wavelets db2 - db20 (even index numbers only) are commonly
used, and db2 is called Haar wavelet. The ‘Haar wavelet’ is the first known wavelet and
was proposed in 1909 by Alfred Haar, and term ‘wavelet’ was comes much later. The
Haar wavelet is a simplest possible wavelet. The disadvantage of the Haar wavelet is that
it is not continuous and therefore not differentiable. The Haar wavelet’s mother wavelet
function ψ(x) can be described as
⎧
⎪
⎨1 0 ≤ x < 1/2,
ψ(x) = −1 1/2 ≤ x < 1,
⎪
⎩
0 otherwise,

and its scaling function φ(x) can be described as

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 4 Mani Mehra and Ratikanta Behera


1 0 ≤ x < 1,
φ(x) =
0 otherwise.

Each wavelet has a number of vanishing moments equal to half the number of co-
efficients. For example db2 has one vanishing moments, db4 has two moments, etc..
A vanishing moment refers to the wavelet ability to represent polynomial behavior or
information in a signal. Mathematically, wavelet is said to have M (M ∈ N) vanishing
moment if its satisfy the following condition

xn ψ(x)dx = 0 for, n = 0, 1, ..., M − 1. (4)

R

Similarly scaling function can represents polynomial of degree up to M − 1 exactly.

This property improves the efficiency of ψ(x) at deliecting singularities in the signal and
represent piecewise smooth function.
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

Daubechies wavelet is widely used in solving a broad range of problems, e.g. convolve
image data, self-likely properties of a signal or fractal problem [3], signal discontinuities
Members Copy, Not for Commercial Sale

and solving partial differential equation [12] etc..

www.IndianJournals.com

————————————————————————————————————
The MATLAB function
[hk , gk ] = wfilters(‘dbM’, ‘iter’) compute these filter coefficients, where M = D2 ,
see [9] for details (e.g.: For Haar wavelet M = 1, D = 2, hk = [ √12 , √12 ] and
gk = [ √12 , − √12 ]).
————————————————————————————————————

One thing is more clear that there is no closed from analytic (explicit) formula for
Daubechies scaling function φ(x) and ψ(x) expect Haar scaling and wavelet function and
its value can be computed only at dyadic point using the cascade algorithm [9, 13].

————————————————————————————————————–
The MATLAB function
[φi , ψi , xi ] =wavefun(‘dbM’, ‘iter’) computes the value of φ(x) and ψ(x) at the grid
xi = [0, 2iter 1
, ..., D − 1], see [13] for details. The functions φ(x) and ψ(x) are plotted in
Figure 1 for D=2 and iter = 4.
—————————————————————————————————————

2.2 Coiflet wavelet

Coiflets [9, 14] are discrete wavelets designed by Ingrid Daubechies, at the request
of Ronald Coifman, to have scaling functions with vanishing moments. These wavelets
are compactly supported with higher number of vanishing moments for both φ and ψ for

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 Applications and Properties of different Kind of Wavelet 5

a given support width. It is an orthogonal wavelet and its order is D. This wavelet is
not symmetric but near symmetric. Coiflet wavelets are mostly used in image processing.

—————————————————————————————————
The MATLAB function
[φi , ψi , xi ] =wavefun(‘coifD’,‘iter’) computes the value of φ(x) and ψ(x), see [14, 9]

coif2: φ coif2:ψ
1.5 2

1
1

0.5

0
0
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

−0.5 −1
0 5 10 15 0 5 10 15
Members Copy, Not for Commercial Sale
www.IndianJournals.com

Figure 2:

for details. The functions φ(x) and ψ(x) are plotted in Figure 2 for D = 2 and
iter=4.
—————————————————————————————————–

Imaginary part of shan1.5−1 Real part of shan1.5−1

1 1

0.5 0.5

0 0

−0.5 −0.5

−1 −1
−20 −10 0 10 20 −20 −10 0 10 20

Figure 3:

2.3 Shannon wavelet

The analytic expression of ψ(x) for shannon wavelet is given by

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 6 Mani Mehra and Ratikanta Behera

sin(πx)
ψ(x) =
(πx)
This wavelet is derived from a function that is proportional to the cosine function.
Moreover, it is non-orthogonal with infinite support and maximum energy occupies wider
bond around origin. It is a family of complex wavelets (i.e. wavelet and scaling functions
are complex functions). Symmetricity and explicit expression of ψ are two advantages
of this wavelet. We can’t make fast algorithms for decomposition and reconstruction
of this wavelet (measure difficulties). Shannon wavelets are mostly used in signal analy-
sis and in particular, for used in the reconstruction of a signal from a digital sampling [1].
—————————————————————————————————
The MATLAB function
[ψi , xi ] = shanwavf(lb, ub, n, fb, fc) computes the real part of ψ(x) and imaginary
part of ψ(x), where fb=1, fc=1.5, lb=-20, ub=20, n=1000, see [11] for details. The
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

function real part and imaginary part of ψ(x) are plotted in Figure 3.
—————————————————————————————————–
Members Copy, Not for Commercial Sale
www.IndianJournals.com

2.4 Meyer wavelet

Meyer wavelet [11] is infinitely regular orthogonal wavelet. Moreover, the ψ(x) and
φ(x) do not have compact support but its fourier transform is compactly supported. It
is symmetric and possible to construct Discrete Wavelet Transform (DWT) but without
Fast Wavelet Transform (FWT) [1, 15]. One of the application of Meyer wavelet is in
regularization [16].
—————————————————————————————————-
The MATLAB function
[φi , ψi , xi ] = meyer(lb, ub, n) computes the value of φ(x), and ψ(x), where lb = -8, ub
= 8 and n = 1024 see [11] for more details. The functions φ(x) and ψ(x) are plotted in
Figure 4.
—————————————————————————————————–
Meyer scaling function Meyer wavelet
1.5 1.5

1
1
0.5
0.5
0
0
−0.5

−0.5 −1
−10 −5 0 5 10 −10 −5 0 5 10

Figure 4:

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 Applications and Properties of different Kind of Wavelet 7

Morlet wavelet Mexican hat wavelet

1 1

0.5
0.5
0
y

y
0
−0.5

−1 −0.5
−4 −2 0 2 4 −5 0 5
x x

Figure 5: (a) Morlet wavelet (b) Maxican Hat wavelet

2.5 Mexican hat wavelet

Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

The analytic expression of ψ(x) for Mexican hat wavelet is given by

Members Copy, Not for Commercial Sale
www.IndianJournals.com

2
ψ(x) = (1 − 2x2 )e−x

Maxican wavelet [9] is the normalized second derivative of Gaussian probability den-
sity function. It is a special cases of the family of continuous wavelets (i.e. continuous
wavelets are functions used by the continuous wavelet transform. These functions are
defined as analytical expressions, as functions either of time or frequency). It is also
called as “hermitian wavelets”. These wavelets are non-orthogonal, with infinite support
and has maximum energy around origin with the narrow band. Symmetry and explicit
expression of ψ(x) are two nice properties of this wavelet. Here also, we can’t make fast
algorithm for decomposition and reconstruction like Shannon wavelet. It is frequently
used as a blob detector and for automatic scale selection in computer vision applications
and also used in image processing [17].

—————————————————————————————————
The MATLAB function
[ψi , xi ] = maxihat(lb, ub, n) computes the value of ψ where lb = -5, ub = 5
and n = 1000 see [9] for more details. The functions ψ(x) is plotted in Figure
5(b).
—————————————————————————————————–
2.6 Morlet wavelet
The analytic expression of ψ(x) for Morlet wavelet is given by
2 /2
ψ(x) = e−x . cos(5x)

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 8 Mani Mehra and Ratikanta Behera

Morlet wavelet [2, 9] is derived from a function that is proportion to the cosine func-
tion and Gaussian probability density function. It is non-orthogonal, infinity support
and maximum energy lies around origin with narrow band. Symmetry and explicit ex-
pression of ψ(x) are again two nice property of this wavelet and difficulty of making fast
algorithms for decomposition and reconstruction remains same. This wavelets are mostly
used in signal analysis. Sometimes it is also used for the detection of ecological patterns
[18].

—————————————————————————————————–
The MATLAB function
[ψi , xi ] = morlet(lb, ub, n) computes the value of ψ where lb = -4, ub = 4 and n =
1000 see [9] for more details. The functions ψ(x) is plotted in Figure 5(a).
—————————————————————————————————–
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

bior2.2: φ1 bior2.2: ψ1
4 4
Members Copy, Not for Commercial Sale
www.IndianJournals.com

3
2
2

1 0

0
−2
−1

−2 −4
0 1 2 3 4 5 0 1 2 3 4 5
bior2.2: φ bior2.2: ψ
1.5 2

1.5

1 1

0.5

0.5 0

−0.5

0 −1
0 1 2 3 4 5 0 1 2 3 4 5

Figure 6:

2.7 Bi-orthogonal wavelet

A biorthogonal wavelet is a wavelet where the associated wavelet transform is in-
vertible but not necessarily orthogonal [9]. In the biorthogonal case, every MRA is

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 Applications and Properties of different Kind of Wavelet 9

accompanied by dual MRA consonsisting of nested V˜j with bases given by dual scaling
function ψ̃kj which are biorthogonal to scaling function

φjk , φ̃jk  = δk,k , for k, k  ∈ I,


and the pair (ψkj , ψ̃kj  ) satisfy the following biorthogonality property

ψkj , ψ̃kj   = δj,j  δk,k , for j, j  , k, k  ∈ I.

if φjk = φ̃jk , ∀j, k ⇒ orthogonal scaling function.

if V j = V˜j , ∀j ⇒ semiorthogonal scaling function.

—————————————————————————————————————
The MATLAB function
[φ̃i , ψ̃i , φi , ψi , xi ] =wavefun(‘bior2.2’,‘iter’) computes the value of φ̃(x), ψ̃(x), ψ(x),
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

and φ(x) at the grid xi = [0, 2iter 1

, ..., D − 1], see [9] for details. The functions
φ̃(x), ψ̃(x), ψ(x), and φ(x) are plotted in Figure 6 for ‘bio2.2’ (both wavelet and dual
Members Copy, Not for Commercial Sale
www.IndianJournals.com

wavelet have vanishing moments) and iter = 4.

—————————————————————————————————————
The Bi-orthogonal wavelet system generalize the classical orthogonal wavelet sys-
tem. It have one additional degree of freedom is the possibility to construct symmetric
wavelet function. Since it is invertible, the perfect reconstruction and symmetric wavelet
functions exist in biorthogonal wavelets, more details see [9]. Biorthogonal wavelets are
mostly used in image compression and solving partial differential equation [19].

2.8 Interpolating wavelet

Interpolating wavelets are independently discovered by Donoho [20] and Harten [21].
Interpolating wavelets are based on the interpolating subdivision scheme of Deslauriers
and Dubuc [22]. The scaling functions {φjk |j ≥ 0, k ∈ V j }, are called interpolating, if a
set of points {xjk |j ≥ 0, k ∈ V j }, with xjk = xj+1
2k exists, so that

∀k, k  ∈ V j : φjk (xjk ) = δk,k .

In the case of interpolating scaling functions, we can always take the dual scaling
functions to be Dirac distributions, φ̃jk = δ(x − xjk ) which is necessarily biorthogonal.
The set of filters resulting from interpolating scaling functions, can be seen as a dual
lifting of the Lazy wavelet [23]. So we have an interpolating multiresolution analysis
which is dually lifted from the Lazy wavelet transform [23] is an orthogonal transform
that simple subsamples the coefficients. The filters of the Lazy wavelet transform are
given as
ajk,l = ãjk,l = δk,l and bjk,l = b̃jk,l = δm,l .

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 10 Mani Mehra and Ratikanta Behera

This wavelet has many advantages such as compact support, linear phase, high regularity,
and finite digit of filters. One of the disadvantage is that, they do not have any vanishing
moments and thus do not form a Riesz basis for L2 (R). This wavelet is mostly used in
solving partial differential equation [24] and in image compression.

2.9 Second Generation wavelet

Second generation wavelets are a generalization of our traditional wavelets, which are
more easily applied to functions defined on domains more general than Rn . The main
advantage of using second generation wavelets is that wavelets can be custom designed for
complex domains and irregular sampling. The construction of second generation wavelet
is done with the help of the lifting scheme [25].
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

Analysis scaling function (phiA) Analysis wavelet function (psiA)

1.5 0.5
Members Copy, Not for Commercial Sale
www.IndianJournals.com

0
1
−0.5
0.5
−1

0 −1.5
−0.5 0 0.5 1 1.5 0 1 2 3
Synthesis scaling function (phiS) Synthesis wavelet function (psiS)
1.5 2

1 1
0.5
0
0
−1
−0.5
0 1 2 3 0 0.5 1

Figure 7:

Lifting scheme is an efficient algorithm to calculate wavelet transforms and it allows

the construction of second generation wavelets [23, 26]. Second generation wavelets are
widely used in solving a broad range of problems, e.g. wavelets on bounded domains,

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 Applications and Properties of different Kind of Wavelet 11

wavelets on curves and surfaces, wavelets and irregular sampling [27] etc.

Lifting scheme has more advantages i.e. It is very easy to learning because does
not rely on the Fourier transform and its allows a faster implementation of the wavelet
transform, and its also allows a fully in-place calculation of the wavelet transform means
its no need extra auxiliary memory and immediately inversible and its works well for a
bounded domain.

—————————————————————————————————————
The MATLAB function
lshaar=liftwave(‘haar’);
els = {‘p’, [-0.125 0.125], 0};
lsnew = addlift(lshaar, els);
[ai , bi , ãi , b˜i ] =ls2filt(lsnew);
bswfun(ai , bi , ãi , b˜i , ‘plot’);
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

where ai , bi , ãi and b˜i are low pass and high pass filter coefficients. The analysis scaling
Members Copy, Not for Commercial Sale

and wavelet function are plotted at the top of Figure 7. The synthesis scaling and wavelet
www.IndianJournals.com

function are plotted in Figure 7 at the bottom.

—————————————————————————————————————

3 Comparison of wavelets
Generally we divide wavelets in four categories in Figure 8. Where the crude wavelet
means “A wavelet is said to be crude when satisfying only the admissibility condition
∞
( −∞ ψ(x)dx = 0, which ensures that the inverse transform and Parseval formula are
applicable)”, see details in [9, 28]. Moreover, Complex wavelet, Infinitely regular wavelet
and compactly supported wavelet are other categories. Then we point out which wavelet
are belongs in which category. Moreover, we also discuss the main advantages and
disadvantage of these different categories wavelets are explained in Table 1. Finally,
we compare the properties of different type of wavelet is in Table 2.

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 12 Mani Mehra and Ratikanta Behera

WAVELETS

CRUDE WAVELET COMPLEX WAVELET INFINITELY REGULAR WAVELET COMPACTLY SUPPORTED WAVELET

MEYER WAVELET ORTHOGONAL WAVELET BIORTHOGONAL WAVELET

SHANNON WAVELET

COMPLEX SHANNON WAVELET

MEXICAN WAVELET DAUBECHIES WAVELET

BíSPLINE WAVELET
COMPLEX MORLET WAVELET
MORLET WAVELET COIFLETS WAVELET

COMPLEX FREQUANCY BíSPLINE WAVELET

Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

Figure 8:
Members Copy, Not for Commercial Sale
www.IndianJournals.com

Category of Wavelets Advantages Disadvantages

Crude wavelet Symmetric and Fast algorithm and

ψ has explicit expression reconstraction is unavaible
Complex wavelet Symmetric Fast algorithm and
ψ has explicit expression reconstraction is unavaible
Infinitely regular wavelet Symmetry, infinite regularity Fast algorithm unavailable

Orthogonal and compactly Compactly supported, Poor regularity

supported wavelet vanishing moments
and FIR filters
Symmetry with FIR filters,
Biorthogonal and desirable properties for Orthogonality is lost
compactly supported decomposition and reconstruction
wavelet are split and nice allocation
is possible

Table 1

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 Applications and Properties of different Kind of Wavelet 13

Property dbM coifN Shan meyr mexh morl bior

Infinity regularity × ×     ×
Arbitary regularity   × × × × 
Compactly supported   × × × × ×
orthogonal wavelet
Compactly supported × × × × × × 
biorthogonal wavelet
Symmetry × ×     
Anti Symmetry  × × × × × ×
Near Symmetry ×  × × × × ×
Arbitary number
of   × × × × 
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

vanishing moment
Vanishing moments ×  × × × × ×
Members Copy, Not for Commercial Sale

of φ
www.IndianJournals.com

Existence of φ   ×  × × 
Orthogonal Analysis   ×  × × ×
Bi-orthogonal analysis   ×  × × 
Extract      ≈ 
Reconstruction
FIR filter   × × × × 
Continues transform   ×    
Discrete transform   ×  × × 
Fast Algorithm   × × × × 
Explicit Expression × ×  ×   for
splines
(dbN = Daubechies wavelet, coifN = Coiflet wavelet, shan = Shannon wavelet, meyr = Meyer wavelet, mexh = Mexican
hat wavelet, morl = Morlet wavelet, bior = Biorthogonal wavelet)

Table 2

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 14 Mani Mehra and Ratikanta Behera

References
[1] S G Mallat. A wavelet tour of signal processing. Academic Press.

[2] A. Grossman P. Goupillaud and J. Morlet. Cycle-octave and related transforms in

seismic signal analysis. Geoexploration, 23:85–102, 1984.

[3] S. T. Welstead. Fractal and wavelet image compression techniques. SPIE Press,
1999.

[4] D.Fohar O. Mella E. Didiot, I. Illina*. A wavelet–based parameterization for

speech/musis discimination. Computer Speech and Language, 24:341–357, 2009.

[5] P. A. Davidson. Turbulence An Introduction for Scientists and Engineers. Oxford

University Press, 2004.
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

[6] M. Farge, K. Schneider, and N. K.-R. Kevlahan. Non-gaussianity and coherent vor-
tex simulation for two-dimensional turbulence using an adaptive orthogonal wavelet
Members Copy, Not for Commercial Sale

basis. Phys. Fluids, 11:2187–2201, 1999.

www.IndianJournals.com

[7] D. Stensrud. Parameterization Schemes: Keys to Understanding Numerical Weather

Prediction Models. Cambridge University Press, 2007.

[8] A. Grossmann and J. Morlet. Decomposition of hardy functions into square intr-
grable wavelets of constant shape. SIAM J.Math. Anal., 15:723–736, 1984.

[9] I. Daubechies. Ten lectures on wavelets. SIAM, 1992.

[10] S.G. Mallat. Multiresolution approximations and wavelet orthonormal bases of lR .

Trans. Amer. Math. Soc., 315(1):69–87, 1989.

[11] Y. Meyer. Analysis at urbana I: Analysis in function spaces. Cambridge University

Press, Cambridge, 1989.

[12] B. V. Rathish Kumar and Mani Mehra. Time-accurate solution of advection-

diffusion problems by wavelet taylor galerkin method. Communications in Numerical
Methods in Engineering, 21:313–326, 2005.

[13] G. Strang and T. Nguyen. wavelets and filter banks. wellesley-cambridge press, 1996.

[14] R. Coifman G. Beylkin and V. Rokhlin. Fast wavelet transforms and numerical
algorithms. Comm. Pure Appl. Math, 44:141–183, 1991.

[15] E. Beckenstein George Bachman, L. Narici. Fourier and Wavelet. Springer.

[16] X. A. Shen and G. G. Walter. Meyer wavelet regularization. Numerical functional

analysis and optimization, 23(1):195–215, 2002.

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013
 Applications and Properties of different Kind of Wavelet 15

[17] H. Guo, Q. He, and B. Jiang. The application of mexican hat wavelet filtering and
averaging algorithm in raman spectra denoising. Proceedings of the 2008 Congress
on Image and Signal Processing, 4:321–326, 2008.

[18] X. Mi., H. Ren, Z. Ouyang, W. Wei, and K. Ma. The use of the maxican hat and the
morlet wavelets for detection of ecological patterns. Plant Ecology, 179:1–90, 2005.

[19] Mani Mehra and Nicholas K.-R. Kevlahan. An adaptive wavelet collocation method
for the solution of partial differential equation on the sphere. Journal of Computa-
tional Physics, 227:5610–5632, 2008.

[20] David L. Donoho. Interpolating wavelet transforms. Technical Report 408, Depart-
ment of Statistics, Stanford University, 1992.

[21] A Harten. Adaptive multiresolution schemes for shock computations. J. Computa-

tional Physics, 115:319–338, 1994.
Downloaded From IP - 180.149.52.43 on dated 4-Oct-2013

[22] G. Deslauriers and S. Dubuc. Symmetric iterative interpolation process. Constr.

Members Copy, Not for Commercial Sale

Appro., 5(49), 1989.

www.IndianJournals.com

[23] W. Sweldens. The lifting scheme: A construction of second generation wavelets.

SIAM J. Math. Anal., 2(29):511–546, 1998.

[24] M. Holmstrom and J. Walden. Adaptive wavelet methods for hyperbolic pde’s. J.
Comput. Phys., 13:19–49, 1998.

[25] W. Sweldens. The lifting scheme: A new philophy in biorthogonal wavelet construc-
tions. Katholieke Universiteit Leuven Belgium, Department of Computer Science.

[26] Wim Sweldens. The lifting scheme: A custom-design construction of biorthogonal

wavelets. Applied and computational harmonic analysis, 3(2):186–200, 1996.

[27] Mani Mehra and Nicholas K.-R. Kevlahan. An adaptive multilevel wavelet solver
for elliptic equations on an optimal spherical geodesic grid. SIAM J Sci. Comput.,
30(6):3073–3086, 2008.

[28] C. K. Chui. An introduction to wavelets. Academic Press, Boston,MA, 1992.

INDIAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS Vol. 4, Issue 1, Jan-June 2013

View publication stats