Article Outline
Glossary
Definition of the Subject
Introduction
Refinable Function Vectors
Multiresolution Approximation and Discrete Multiwavelet Transform
Moments and Approximation Order
The Discrete Multiwavelet Transform (DMWT) Algorithm
Pre- and Postprocessing and Balanced Multiwavelets
Boundary Handling
Applications
Polyphase Factorization
Lifting
Two‐Scale Similarity Transform (TST)
Examples and Software
Future Directions
Bibliography
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Abbreviations
- Balanced multiwavelet:
-
A multiwavelet for which the expansion coefficients for polynomials up to a certain degree are polynomial sequences. Balanced multiwavelets do not require pre- or postprocessing.
- Discrete multiwavelet transform (DMWT):
-
The algorithm which decomposes a signal into a coarse approximation and fine detail at several levels. A direct generalization of the Discrete Wavelet Transform (DWT) for scalar wavelets.
- Lifting:
-
A method for modifying an existing multiwavelet, or building one from scratch. Lifting can be used to impose approximation order, balancing, or symmetry.
- Modulation matrix:
-
The modulation matrix
$$ M(\xi) = \begin{pmatrix} H(\xi) & H(\xi+\pi) \\ G(\xi) & G(\xi+\pi) \end{pmatrix} \:, $$can be used to describe the action of the DMWT. It is also used in the construction of multiwavelets by lifting or TST.
- Multiresolution, multiresolution approximation (MRA):
-
Multiresolution is the fundamental concept underlying everything related to any kind of wavelet. A signal is decomposed into a low resolution approximation, plus fine detail at one or more levels of resolution. An MRA is a nested chain of subspaces of L 2 which describes this concept mathematically.
- Multiscaling function, multiwavelet function,wavelet:
-
A multiwavelet of multiplicity r consists of a multiscaling function and a multiwavelet function. Both are functions from ℝ to ℂr, that is, function vectors. Multiwavelets are not the same as multivariate wavelets, which are functions from \( { \mathbb{R}^{n} } \) to ℂ.
- Polyphase decomposition, polyphase matrix:
-
The polyphase decomposition separates a coefficient sequence into two sequences, by even and odd subscripts. It can be used to describe or implement the DMWT.The polyphase matrix
$$ P(z) = \begin{pmatrix} H^{(0)}(z) & H^{(1)}(z) \\ G^{(0)}(z) & G^{(1)}(z) \end{pmatrix}\:, $$is useful in the construction of multiwavelets, especially orthogonal multiwavelets.
- Pre- and postprocessing:
-
Before a DMWT can be applied to a signal, the signal needs to be written as a multiscaling function series. The DMWT works on the coefficients of the series expansion, not on the values of the signal.For scalar wavelets the distinction can usually be ignored, but not for multiwavelets. Having a multiwavelet of approximation order p does not mean that the coefficients of a polynomial up to order \( { p-1 } \) form a polynomial sequence.Preprocessing converts point samples into expansion coefficients. Postprocessing does the opposite.
- Refinable function vector, refinement equation:
-
A refinable function vector of multiplicity r is a function from ℝ to \( { \mathbb{C}^{r} } \) which satisfies a refinement equation of the form
$$ \boldsymbol{\phi}(x) = \sqrt{2} \sum_{k=k_{0}}^{k_{1}} H_{k} \, \boldsymbol{\phi}(2x-k) $$with recursion coefficients H k which are \( { r\! \times\! r } \) matrices.
- Symbol:
-
Given any sequence \( { \mathbf{a} = \{a_{k} \} } \), the symbol of \( { \mathbf{a} } \) is defined as
$$ \begin{aligned} a(\xi) &= \sum_{k} a_{k} \text{e}^{-i\xi} \quad \text{or} \\[1mm] a(z) &= \sum_{k} a_{k} z^{k}\:,\quad z = \text{e}^{-i\xi}\:, \end{aligned}$$possibly with a normalizing factor. The sequence may represent point samples of a signal, the recursion coefficients of a two‐scale refinement equation, or other quantities.Both the trigonometric and polynomial notations are useful, depending on the setting; it is trivial to switch back and forth between them.
- Two‐scale similarity transform (TST):
-
The TST is a way of moving approximation orders in a biorthogonal multiwavelet pair from one side to the other. For scalar wavelets it corresponds to moving a factor of \( { (1+\text{e}^{-i\xi})/2 } \) back and forth.The TST can be used to modify an existing multiwavelet, or to build multiwavelets from scratch. It can also be used to impose or characterize symmetry, approximation order and balancing.
Bibliography
Alpert BK (1990) Sparse representation of smooth linear operators.TechnicalReport YALEU/DCS/RR-814. Yale University, New Haven
Alpert BK, Beylkin G, Coifman R, Rokhlin V (1990) Wavelets for the fast solutionof second‐kind integral equations.Technical Report YALEU/DCS/RR-837. Yale University, New Haven
Alpert B, Beylkin G, Gines D, Vozovoi L (2002) Adaptive solution of partialdifferential equations in multiwavelet bases.J Comput Phys 182(1):149–190
Alpert BK (1993) A class of bases in L 2 for the sparse representation of integral operators.SIAM J Math Anal24(1):246–262
Andersson L, Hall N, Jawerth B, Peters G (1994) Wavelets on closed subsets of thereal line.In: Recent advances in wavelet analysis. Academic Press, Boston, pp 1–61
Attakitmongcol K, Hardin DP, Wilkes DM (2001) Multiwavelet prefilters II: optimalorthogonal prefilters.IEEE Trans Image Proc 10(10):1476–1487
Averbuch AZ, Zheludev VA (2002) Lifting scheme for biorthogonal multiwaveletsoriginated from Hermite splines.IEEE Trans Signal Process 50(3):487–500
Averbuch A, Israeli M, Vozovoi L (1999) Solution of time-dependent diffusionequations with variable coefficients using multiwavelets.J Comput Phys 150(2):394–424
Averbuch A, Braverman E, Israeli M (2000) Parallel adaptive solution ofa Poisson equation with multiwavelets.SIAM J Sci Comput 22(3):1053–1086
Bacchelli S, Cotronei M, Lazzaro D (2000) An algebraic constructionof k‑balanced multiwavelets via the lifting scheme.Numer Algorithms23(4):329–356
Bacchelli S, Cotronei M, Sauer T (2002) Multifilters with and withoutprefilters.BIT 42(2):231–261
Bramble JH, Cohen A, Dahmen W, Canuto C (ed) (2003) Multiscale problems andmethods in numerical simulations. Lecture Notes in Mathematics, vol 1825. Springer, Berlin.Lectures given at the CIME Summer School held inMartina Franca, 9–15 September, 2001
Brislawn C (1996) Classification of nonexpansive symmetric extension transformsfor multirate filter banks.Appl Comput Harmon Anal 3(4):337–357
Bui TD, Chen G (1998) Translation-invariant denoising using multiwavelets.IEEE Trans Signal Process 46(12):3414–3420
Burrus CS, Gopinath RA, Guo H (1998) Introduction to wavelets and wavelettransforms: a primer.Prentice Hall, New York
Chen Z, Micchelli CA, Xu Y (1997) The Petrov–Galerkin method for secondkind integral equations II: multiwavelet schemes.Adv Comput Math 7(3):199–233
Cohen A, Daubechies I, Vial P (1993) Wavelets on the interval and fast wavelet transforms. Appl Comput Harmon Anal 1(1):54–81
DahmenW, Han B, Jia RQ, Kunoth A (2000) Biorthogonal multiwavelets on theinterval: cubic Hermite splines.Constr Approx 16(2):221–259
Daubechies I, Sweldens W (1998) Factoring wavelet transforms into liftingsteps.J Fourier Anal Appl 4(3):247–269
Davis GM, Strela V, and Turcajova R (2000)Multiwavelet construction via the lifting scheme.In: Wavelet analysis and multiresolution methods. Dekker, New York, pp57–79
Donovan GC, Geronimo JS, Hardin DP, Massopust PR (1996) Construction oforthogonal wavelets using fractal interpolation functions.SIAM J Math Anal 27(4):1158–1192
Downie TR, Silverman BW (1998) The discrete multiple wavelet transform andthresholding methods.IEEE Trans Signal Process 46(9):2558–2561
Efromovich S (2001) Multiwavelets and signal denoising.Sankhyā Ser A63(3):367–393
Goh SS, Yap VB (1998) Matrix extension and biorthogonal multiwaveletconstruction.Linear Algebr Appl 269:139–157
Goh SS, Jiang Q, Xia T (2000) Construction of biorthogonal multiwavelets usingthe lifting scheme.Appl Comput Harmon Anal 9(3):336–352
Han B, Jiang Q (2002) Multiwavelets on the interval.Appl Comput Harmon Anal12(1):100–127
Hardin DP, Marasovich JA (1999) Biorthogonal multiwavelets on \( { [-1,1] } \).Appl Comput Harmon Anal7(1):34–53
Hardin DP, Roach DW (1998) Multiwavelet prefilters I: orthogonal prefilterspreserving approximation order \( { p \le 2 } \).IEEE Trans Circuits Syst II. Analog Digital Signal Process 45(8):1106–1112
Heil C, Strang G, Strela V (1996) Approximation by translates of refinablefunctions.Numer Math 73(1):75–94
Hsung TC, Lun DPK, Ho KC (2005) Optimizing the multiwavelet shrinkagedenoising.IEEE Trans Signal Process 53(1):240–251
Johnson BR (2000) Multiwavelet moments and projection prefilters.IEEE TransSignal Process 48(11):3100–3108
Kautsky J, Turcajova R (1995) Discrete biorthogonal wavelet transforms as blockcirculant matrices.Linear Algebr Appl 223/224:393–413
Keinert F (2001) Raising multiwavelet approximation order through lifting.SIAM J Math Anal 32(5):1032–1049
Keinert F (2004) Wavelets and multiwavelets.Studiesin Advanced Mathematics. Chapman & Hall/CRC, Boca Raton
Kessler B (2005) Balanced scaling vectors using linear combinations of existingscaling vectors.In: Approximation theory, XI. Mod Methods Math. Nashboro Press, Brentwood, pp 197–210
Lawton W, Lee SL, Shen Z (1996) An algorithm for matrix extension and waveletconstruction.Math Comp 65(214):723–737
Lebrun J, Vetterli M (2001) High-order balanced multiwavelets: theory,factorization, and design.IEEE Trans Signal Process 49(9):1918–1930
Lian JA (2005) Armlets and balanced multiwavelets: flipping filterconstruction.IEEE Trans Signal Process 53(5):1754–1767
Lin EB, Xiao Z (2000) Multiwavelet solutions for the Dirichlet problem.In:Wavelet analysis and multiresolution methods. Lecture Notes in Pure and Appl Math, vol 212. Dekker, New York, pp241–254
Madych WR (1997) Finite orthogonal transforms and multiresolution analyses onintervals.J Fourier Anal Appl 3(3):257–294
Maleknejad K, Yousefi M (2006) Numerical solution of the integral equation ofthe second kind by using wavelet bases of Hermite cubic splines.Appl Math Comput 183(1):134–141
Massopust PR (1998) A multiwavelet based on piecewise C 1 fractal functions and related applications to differential equations.Bol Soc Mat Mex4(2):249–283
Micchelli CA, Xu Y (1994) Using the matrix refinement equation for theconstruction of wavelets II: smooth wavelets on \( { [0,1] } \).In: Approximation and computation. Birkhäuser, Boston, pp 435–457
Micchelli CA, Xu Y (1994) Using the matrix refinement equation for theconstruction of wavelets on invariant sets.Appl Comput Harmon Anal 1(4):391–401
Miller J, Li CC (1998) Adaptive multiwavelet initialization.IEEE Trans SignalProcess 46(12):3282–3291
Plonka G (1997) Approximation order provided by refinable function vectors.Constr Approx 13(2):221–244
Plonka G, Strela V (1998) Construction of multiscaling functions withapproximation and symmetry.SIAM J Math Anal 29(2):481–510
Plonka G, Strela V (1998) From wavelets to multiwavelets.In: Mathematicalmethods for curves and surfaces, II. Vanderbilt Univ Press, Nashville, pp 375–399
Resnikoff HL, Tian J, Wells RO Jr (2001) Biorthogonal wavelet space:parametrization and factorization.SIAM J Math Anal 33(1):194–215
Selesnick IW (1998) Multiwavelet bases with extra approximation properties.IEEE Trans Signal Process 46(11):2898–2908
Selesnick IW (2000) Balanced multiwavelet bases based on symmetric FIR filters.IEEE Trans Signal Process 48(1):184–191
Shen L, Tan HH (2001) On a family of orthonormal scalar wavelets andrelated balanced multiwavelets.IEEE Trans Signal Process 49(7):1447–1453
Shi H, Cai Y, Qiu Z (2006) On design of multiwavelet prefilters.Appl MathComput 172(2):1175–1187
Strang G, Nguyen T (1996) Wavelets and filter banks.Wellesley-CambridgePress, Wellesley
Strela V (1996) Multiwavelets: theory and applications.Ph D thesis,Massachusetts Institute of Technology
Strela V, Heller PN, Strang G, Topiwala P, Heil C (1999) The application ofmultiwavelet filter banks to image processing.IEEE Trans Signal Process 8:548–563
Tausch J (2002) Multiwavelets for geometrically complicated domains and theirapplication to boundary element methods.In: Integral methods in science and engineering. Birkhäuser, Boston, pp251–256
Tham JY, Shen L, Lee SL, Tan HH (2000) A general approach for analysis andapplication of discrete multiwavelet transform.IEEE Trans Signal Process 48(2):457–464
Tham JY (2002) Multiwavelets and scalable video compression.Ph D thesis,National University of Singapore. Available at http://www.cwaip.nus.edu.sg/thamjy
Turcajová R (1999) Construction of symmetric biorthogonal multiwavelets bylifting. In: Unser MA, Aldroubi A, Laine AF (eds) Waveletapplications in signal and image processing (SPIE), VII, vol 3813. SPIE, Bellingham, pp 443–454
von Petersdorff T, Schwab C, Schneider R (1997) Multiwavelets forsecond‐kind integral equations.SIAM J Numer Anal 34(6):2212–2227
VrhelMJ, Aldroubi A (1998) Projection based prefiltering for multiwavelet transforms.IEEE Trans Signal Process 46(11):3088
Xia T, Jiang Q (1999) Optimal multifilter banks: design, related symmetricextension transform, and application to image compression.IEEE Trans Signal Process 47(7):1878–1889
Xia XG, Geronimo JS, Hardin DP, Suter BW (1996) Design of prefilters fordiscrete multiwavelet transforms.IEEE Trans Signal Process 44(1):25–35
Xia XG (1998) A new prefilter design for discrete multiwavelet transforms.IEEE Trans Signal Process 46(6):1558–1570
Yang S, Peng L (2006) Construction of high order balanced multiscalingfunctions via PTST.Sci Chin Ser F 49(4):504–515
Yang S, Wang H (2006) High-order balanced multiwavelets with dilationfactor a.Appl Math Comput 181(1):362–369
Zhang JK, Davidson TN, Luo ZQ, Wong KM (2001) Design of interpolatingbiorthogonal multiwavelet systems with compact support.Appl Comput Harmon Anal 11(3):420–438
Zhou S, Gao X (2006) A study of orthonormal multi-wavelets on the interval.Int J Comput Math 83(11):819–837
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer-Verlag
About this entry
Cite this entry
Keinert, F. (2012). Multiwavelets . In: Meyers, R. (eds) Computational Complexity. Springer, New York, NY. https://doi.org/10.1007/978-1-4614-1800-9_127
Download citation
DOI: https://doi.org/10.1007/978-1-4614-1800-9_127
Publisher Name: Springer, New York, NY
Print ISBN: 978-1-4614-1799-6
Online ISBN: 978-1-4614-1800-9
eBook Packages: Computer ScienceReference Module Computer Science and Engineering