Professional Documents
Culture Documents
Department of Electrical and Computer Engineering, Rice University, Houston, TX 77251–1892, USA
cient is given as
! "$#&%('*) /
, where 021 is the variance of L
3
[1, 4]. Because
" ’s# %(are
'+)+,.- unknown,
# M 16
01
4 ?@BADC$
E 1
0 1 J
we use estimated values
5
instead, and we obtain an empirical Wiener filter given as (2)
4 '*NO
! 4 16
/
$7
M
L 4 1
E
1
0 1 R 4 1
1
(1)
1
0 1 ?FHG IP
E5J5Q 7
1
0 1TS 1
0 1 U
This work was supported by the National Science Foundation, grant
no. MIP–9457438, DARPA, grant no. DARPA/AFOSR F49620–97–1–
'*NHO (3)
The 4 term ? @BA$C represents the MSE of the Wiener filter when
0513, and Texas Instruments.
Email: choi@ece.rice.edu, richb@rice.edu
Web: http://www.dsp.rice.edu we have
VW
for all ; ?FHG I is the error resulting from
the mismatch of the signal model to the true signal. Once we fix 4. ANALYSIS OF THE WIENERCHOP ALGORITHM
the wavelet transform, ? @BADC is determined for a given signal. We
can see that ? @BADC attains its minimum when the signal energy is
concentrated at a single wavelet coefficient
YXZ for some X and Although wavelet domain denoising using hard thresholding can
5
[]\ for V_
^ X . In other words, the we obtain a good per- be used to estimate a signal, this estimated signal can also be used
formance of the optimal Wiener filter when the wavelet transform as a signal model to design a filter as in (1). Then, we can filter
compacts the signal well. the original signal using this filter to obtain a better estimate of the
The term ? FHG I
EKJ Q represents the error resulting from mis-
signal, because the signal model used to design the filter (obtained
by hard thresholding the wavelet coefficients in :9 ) may be bet-
match of the signal model Q with the signal . When the signal ter matched to the signal than the model in (4). However, using
model is perfect, ? FHG I `\ and we obtain the performance of the the signal model obtained by hard thresholding in 9 to design
optimal Wiener filter. However, we do not know the exact values a Wiener filter in the same domain experiences similar problems
of 5
’s, and this4 error term is not zero. To see how this error (large ? FHG I ) as simple hard thresholding does. Rather, we can
term varies as
changes, consider 4 the behavior of each term think of Wiener filtering in an alternate domain by an orthonormal
in the expression for ? FHG I
E5J 6 in (3). Figure 2 shows the plot transformation.
of4 a term in the summation for 0a\K7(b , 5
dce \57 fZJg\57 fih , and
5
dcje \K7 fKJ$\57 fih . ? FHG I is summation of k suchS terms.4 In this
The WienerChop algorithm in [5] follows from this idea. The
1
0 1 \ otherwise. nate transformation, rendering a more favorable signal model so
that the Wiener filter designed in the new domain can have sig-
4 (4) nificant reduction of error. Then, the noisy signal and the signal
In other words,
in (4) is the model of the signal used to design model in the new domain are given as l 9 , and
the filter, and under this model, the designed Wiener filter is the 2ljnTVn 9 . We can design a Wiener filter in the
hard thresholding algorithm. new domain based on the signal model , and then the filter can
To see how much error would result from hard thresholding, be applied to . We obtain the final time domain signal by inverse
we use the formulas for MSE derived in Section 2. The term ? @BADC transforming back to the time domain.
is fixed once the underlying wavelet transform is determined. In When choosing , we should be careful that the resulting
view of Fig. 4 2, we see that ? FHG I mainly consists of the error result- transform of the signal in the new domain ( ) is as compact as
ing from
z{ ^ 5
for those wavelet coefficients w
rw6|y and possible, so that the inherent Wiener filtering error ( ? @BA$C ) is small.
w /
rw}~y . Such
4 coefficients are removed by hard thresholding, In this respect, it is desirable to choose so that the resulting new
and we have
\ , making the corresponding error term in domain is another wavelet domain. Let 8 denote this wavelet
?FHG I (see (3)) large. Hard thresholding has a poor performance in 9
domain. Then, has the form 86 9 . In terms of 8 ,
the signal is represented as 8 8 in the 8 domain, and
this sense.
9
we have the signal model given by 2 8 8 W9 9 . This
corresponds to having :9 ^ 8 in the algorithm of [5]. When
Other wavelet thresholding or shrinkage algorithms can be
interpreted as Wiener filters in a similar way. And, we expect sim-
ilar behavior of ? FHG I in many other algorithms including the soft 9 ^ 8 , the error ? FHG I of the Wiener filtering in 9 8 domain
depends on the ability of the transform 8 9 to spread
the model mismatches in :9 domain by transforming into 8
thresholding scheme where small wavelet coefficients are also re-
moved. In [1, pp. 425-464], a similar comparison between hard
domain, to make the model suitable for designing a Wiener filter
in 8 domain, assuming the signal is compactly represented in
thresholding and the optimal Wiener filtering (ideal attenuation of
wavelet coefficients) was made by analyzing the errors involved in
each algorithm. 8 as well as in :9 .
In view of the idea of mitigation of signal modeling error to 6. CONCLUSIONS
reduce ? FHG I , we can pick a good pair of wavelet transforms 9
and 8 for a given signal. However, because the original signal In this paper, we analyzed the errors involved in wavelet domain
is unknown, it is hard to characterize the influence of these wavelet empirical Wiener filter. We showed that the error due to the mis-
bases on the estimation error. match of signal model can be significant depending on the methods
of obtaining the signal model. In addition, we noted that this er-
When we have more than two wavelet bases under which the ror can be reduced by transforming both the signal and model to
signal has compact representations, we can consider an iterative another wavelet domain where Wiener filter can be designed with
scheme by choosing a pair of wavelets at a time. The signal esti- smaller error.
mate out of the Wiener filter can again be used as a signal estimate
to design yet another Wiener filter in different wavelet domain. The main contributions of this paper are: (1) Interpretation
However, because the estimation error of the empirical Wiener fil- of hard and soft thresholding denoising schemes as Wiener filters,
tering varies very nonlinearly as the signal model changes, we are (2) Analysis and understanding of WienerChop algorithm of [5],
not guaranteed to have an improvement in performance. Our anal- and (3) Analysis of the errors in empirical Wiener filters, leading
ysis has shown that we can obtain improvement in estimation er- to proper design methods of wavelet domain Wiener filters.
ror compared with hard thresholding when we use the hard thresh- The difficulty in analysis of the wavelet domain Wiener fil-
olded signal estimate to design the Wiener filter. Thus, iterating the tering arises from the signal dependence of the processing. The in-
empirical Wiener filtering for multiple wavelet bases is not guar- fluence of the choice of 9 and 8 on the overall performance
anteed to converge to a good signal estimate. is hard to analyze in general. The development of an algorithm to
design the wavelet transforms for a given signal remains as future
research work.
We have recently learned about another method for denoising
5. DENOISING EXAMPLE using multiple wavelet domains [6]. Currently, we are also investi-
gating filtering algorithms to incorporate more than two wavelet
This idea of mitigating the effect of unfavorable model errors by an transforms to obtain improved performance and the analysis of
orthonormal transformation is well illustrated by an example. Fig- their asymptotic behavior.
ure 3 shows the original noiseless signal obtained by concatenating
Donoho’s Doppler and Blocks signals [3] and the signal corrupted
by white Gaussian noise with variance 0\57(b . Figure 4 shows
7. REFERENCES
the estimated signals using wavelet domain denoising obtained by
hard thresholding in
[1] S. Mallat, A Wavelet Tour of Signal Processing, Academic
1 domain and empirical Wiener filtering.
For the signal by the empirical Wiener filtering in Fig. 4, we chose
Press, 1998.
:9 to be Haar wavelet ( 1 ) and 8 to be Daubechies length-12 [2] D. L. Donoho and I. Johnstone, “Adapting to unknown
wavelet ( as wavelet bases. In terms of 9 and
O 1 ), respectively,
8 , we expect the compaction of wavelet coefficients of our test
smoothness via wavelet shrinkage,” J. Amer. Stat. Assoc.
pp.1200-1224, Dec. 1995.
signal will be almost same because each of these wavelet bases [3] D. L. Donoho, “De-noising by soft-thresholding,” IEEE
compacts half of the signal very well while the compaction of the Trans. Inform. Theory, vol. 41, no. 3, pp. 613-627, May 1995.
other half is not good. Table 1 shows the error term ? @BADC for dif-
ferent choices of wavelet bases. We note that the values of ?@BADC [4] H. V. Poor, An Introduction to Signal Detection and Estima-
are similar for both
1 and O 1 bases, because the compaction
tion, Springer Verlag, 1988.
of wavelet coefficients is almost the same in either domains. To [5] S. P. Ghael, A. M. Sayeed and R. G. Baraniuk, “Improved
see how the estimates of wavelet coefficients and the actual coeffi- Wavelet Denoising via Empirical Wiener Filtering,” Pro-
cients distribute, we computed d Q 4 9 and m 9 . Fig- ceedings of SPIE, vol. 3169, pp. 389-399, San Diego, July
ure 5 show the distribution of
E
$J
B for b6JrqDq&qrJPk . We 1997.
4 coefficients with small magnitudes distribute around
see that the [6] P. Ishwar, K. Ratakonda, P. Moulin, and N. Ahuja, “Image
the line \ , which results in large ?FHG I according 4 to the plot denoising using multiple compaction domains,” Proceedings
Fig. 2. Other coefficients are gathered along the line d . of ICASSP ’98, pp. 1889-1892, Seattle, 1998.
9
To see the distribution of the coefficients and estimates af-
ter transforming into 8 domain, we computed 86 9 Q and
8 and plotted the result in Fig. 6. In this figure4 we see that
there are much less points falling around the line ]\ , result-
ing in much less ?FHG I . This is because the model mismatches in
9 domain, which distributed very unfavorably, was transformed Table 1: Change of error terms depending on the choice of wavelet
to 8 domain and they distribute much more favorably in the bases 9 and 8 . Choosing 9 ` ^ 8 reduces ? FHG I signifi-
8 domain. By actually computing ?FHG I with the model ob- cantly.
tained in the 9 domain by hard thresholding and the signal in 9 1 O1 O1
the same domain, we obtain the values shown in Table 1 for differ- 8 1 O1 O11 1
ent choices of wavelet bases. We observe a significant reduction ?@BADC 2.2165 2.0610 2.0610 2.2165
in ? FHG I when we choose 9 ^ 8 . This reduction in ? FHG I ?FH G I
explains the superiority of the algorithm in [5] over simple hard ?
3.4443
5.6608
2.6966
4.7576
0.7748
2.8358
0.5843
2.8008
thresholding algorithm.
Hard Thresholding
2
y yh -1 sh
W1 H W1 1
0
s −1
W2
−2
^
yhk(=θ) 0 200 400 600 800 1000
^ WienerChop
yk -1 s 2
W2 Hw W2
1
0.2
0.5
0.1
0 0
0.5
0.5
0 −0.5
θ(i) (true wavelet coefficients) 0
^θ(i) (modeled wavelet coefficients)
−0.5 −0.5
^θ
−1
Figure 2: Error due to model mismatch.
−1.5
−1 −0.5 0 0.5 1
θ (true wavelet coefficients)
1
1.5
0
(modeled wavelet coefficients)
−1 1
−2
0 200 400 600 800 1000 0.5
Noisy Signal
2
0
1
−0.5
0
−1
^θ
−1
−2
0 200 400 600 800 1000
−1.5
−1 −0.5 0 0.5 1
Figure 3: Above: DoppleBlock test signal. Below: signal cor- θ (true wavelet coefficients)
rupted by WGN with 0=\57(b .
Figure 6: Model mismatch in the 8 domain.