Abstract
Time domain and frequency domain analysis suffer the drawback of a scarce resolution, in frequency and in time, respectively. In several applications a segmented resolution (multiresolution) is required, where relevant parts of the signal are detailed, and other parts are only roughly represented. The modern technique that achieves this goal is provided by wavelets that are the target of this chapter. Wavelets are strongly related to the topics developed in the previous chapter, since they can be effectively formulated in the framework of generalized transforms (wavelet transform) and their practical implementation is by means of filter banks. In the final part of the chapter, multiresolution analysis is illustrated with examples of applications.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
In the notation \(\varphi_{u}^{(m)}(t)\) the subscript m gives the resolution, and u the shift amount with respect to ϕ (m)(t). In the literature the notation \(\varphi_{n}^{(m)}(t)\), or ϕ mn (t), is used to denote a shifted version of n2m with respect \(\varphi_{0}^{(m)}(t)\).
- 2.
We recall that in our theory “symmetry” is a subspace consisting of “symmetric” signals. Hence, V m and W m are symmetries consisting of signals that have the properties \(s_{m}=\mathcal{P}_{m}[s_{m}]\in V_{m}\) and \(w_{m}=\mathcal{R}_{m}[s_{m}]\in W_{m}\), respectively.
- 3.
- 4.
For the terms and acronyms (nonstandard) we use, see the introduction to this chapter.
- 5.
The author wants to thank Roberto Rinaldo for his contribution to this section.
- 6.
Note that, since we are dealing with separable 2D filters h ij (m,n)=h i (m)h j (n), one can first process the rows and then the columns of the image, without changing the result.
References
E. Candés, L. Demanet, D. Donoho, L. Ying, Fast discrete curvelet transforms. Multiscale Model. Simul. 5, 861–899 (2006)
T. Colthurst, Multidimensional wavelets. Ph.D. Thesis, Massachusetts Institute of Technology, Dept. of Mathematics, June 1997
I. Daubechies, Orthonormal bases of compactly supported wavelets. Commun. Pure Appl. Math. 41, 909–996 (1988)
I. Daubechies, Ten Lecture on Wavelets (SIAM, Philadelphia, 1992)
D. Gabor, Theory of communications. J. Inst. Electr. Eng. 93, 429–457 (1946)
G. Karlsson, M. Vetterli, Extension of finite length signals for sub-band coding. Signal Process. 17, 161–168 (1989)
S. Mallat, A theory of multiresolution signal decomposition: the wavelet representation. IEEE Trans. Pattern Anal. Mach. Intell. 11, 674–693 (1989)
J. Ma, G. Plonka, The curvelet transform: a review of recent applications. IEEE Signal Process. Mag. 27, 118–132 (2010)
R.A. Mayer, C.S. Burrus, A unified analysis of multirate and periodically time varying digital filters. IEEE Trans. Circuits Syst. CAS–29, 327–331 (1982)
A. Said, W.A. Pearlman, A new fast and efficient image codec based on set partitioning in hierarchical trees. IEEE Trans. Circuits Syst. Video Technol. 6, 243–250 (1996)
J. Shapiro, Embedded image coding using zerotrees of wavelet coefficients. IEEE Trans. Signal Process. 41, 3445–3462 (1993)
D. Taubman, M.W. Marcellin, JPEG2000, Image Compression Fundamentals and Practice (Kluwer Academic, Boston, 2002)
P.P. Vaidyanathan, Multirate Systems and Filter Banks (Prentice Hall, Englewood Cliffs, 1993)
M. Vetterli, J. Kovac̆ević, Wavelets and Subband Coding (Prentice Hall, Englewood Cliffs, 1995)
G.K. Wallace, The JPEG still picture compression standard. IEEE Trans. Consum. Electron. 38, 18–34 (1993)
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A: Proof of Proposition 15.7 on Coefficients at Steps m and m+1
We first establish the following statement:
Proposition 15.14
A ℤ m →ℝ interpolator followed by an ℝ→ℤ m+1 decimator is equivalent to a ℤ m →ℤ m+1 decimator. To calculate the global impulse response \(\tilde{g}_{12}(t)\), first calculate the convolution of the component impulse response g 12(t)=g 2(t)∗g 1(t), t∈ℝ, and then apply the ℝ→ℤ m down-sampling, that is,
The statement, illustrated in Fig. 15.38, is a consequence of the general result on the cascade of QIL tf, developed in Chap. 7, as soon as we note that ℤ m ⊃ℤ m+1 (see Problem 15.6). The conclusion is that the combination of the two blocks, each one operating on continuous times, gives a block operating on discrete times, which is the crucial point for the full discretization.
We now prove Proposition 15.7. The link S m (u)→S m+1(v) is given by ℤ m →ℝ interpolator followed by a decimator with impulse response respectively
Then, we apply Proposition 15.14 with g 1=c m+1 and g 1=e m and evaluate the continuous convolution
where, by (15.22),
Hence,
where we have make the variable change τ→2−m τ. Next, we use the two-scale relation (15.46a) with t→(τ−2−m t)/2 and obtain
where in general we cannot use the orthogonality condition. But, restricting t to ℤ m , that is, t=2m k, the integral gives δ n−k . Hence,
To get the impulse response of the link S m (u)→D m+1(v), it is sufficient to replace c m+1 with d m+1 in (15.93) and the two-scale equation (15.46a) with (15.47). The result is \(\tilde{g}_{12}(2^{m}k)=2^{-m}g_{1}^{\ast}(-k)\).
Appendix B: Interpretation of the Expansion of ϕ(t) and ψ(t)
We consider the expansion of the functions ϕ(t) and ψ(t) obtained with the bases \(\boldsymbol{\varPhi}_{-1}=\{\sqrt{2} \varphi (2t-n)\mid n\in\mathbb{Z}\}\)
where the right-hand sides represent the two-scale equations.
The system interpretation of these relations shown in Fig. 15.39. With the introduction of the auxiliary signals
the first of (15.95) can be written as
Then, starting from ϕ(t), g 0(n) is obtained with scale change of \(a=\frac{1}{2}\) followed by an ℝ→ℤ decimator with impulse response q(t), t∈ℝ. In the inverse relation (15.46a), (15.46b) the coefficients g 0(n), n∈ℤ, are ℤ→ℝ interpolated with impulse response \(\sqrt{2} \varphi(t)\) and give the intermediate signal y(t); then, y(t) is dilated by \(a=\frac{1}{2}\) to give the scaling function ϕ(t), t∈ℝ.
A similar interpretation, shown in Fig. 15.39, holds for the generation of the coefficients g 1(n) from the mother wavelet ψ(t) and the reconstruction of ψ(t) from g 1(n).
Using the rules of the Fourier analysis of interpolators and decimators, we find
which give (15.21).
We now prove Proposition 15.11. Combination of (15.63) and (15.66) gives
Next, considering that G 0(f) has period 1 and using again (15.66), we have
which is equivalent to (15.67). In fact, from (15.21) we obtain \(2 |\varPhi(f)|^{2}=|\varPhi(\frac{1}{2}f)|^{2} |G_{0}(\frac{1}{2}f)|^{2}\). Taking the periodic repetition and using (15.66), we get (15.67).
Rights and permissions
Copyright information
© 2011 Springer-Verlag London Limited
About this chapter
Cite this chapter
Cariolaro, G. (2011). Multiresolution and Wavelets. In: Unified Signal Theory. Springer, London. https://doi.org/10.1007/978-0-85729-464-7_15
Download citation
DOI: https://doi.org/10.1007/978-0-85729-464-7_15
Publisher Name: Springer, London
Print ISBN: 978-0-85729-463-0
Online ISBN: 978-0-85729-464-7
eBook Packages: EngineeringEngineering (R0)