Main Content

Wavelet image denoising

denoises the grayscale or RGB image `IMDEN`

= wdenoise2(`IM`

)`IM`

using an empirical Bayesian
method. The `bior4.4`

wavelet is used with a posterior median threshold
rule. Denoising is down to the minimum of `floor(log2([M N]))`

and
`wmaxlev([M N],'bior4.4')`

, where `M`

and
`N`

are the row and column sizes of the image.
`IMDEN`

is the denoised version of `IM`

.

For RGB images, by default, `wdenoise2`

projects the image onto its
principal component analysis (PCA) color space before denoising. To denoise an RGB image
in the original color space, use the `ColorSpace`

name-value
pair.

`[`

returns the scaling and denoised wavelet coefficients in `IMDEN`

,`DENOISEDCFS`

] = wdenoise2(___)`DENOISEDCFS`

using any of the preceding syntaxes.

`[`

returns the scaling and wavelet coefficients of the input image in
`IMDEN`

,`DENOISEDCFS`

,`ORIGCFS`

] = wdenoise2(___)`ORIGCFS`

using any of the preceding syntaxes.

`[`

returns the sizes of the approximation coefficients at the coarsest scale along with the
sizes of the wavelet coefficients at all scales. `IMDEN`

,`DENOISEDCFS`

,`ORIGCFS`

,`S`

] = wdenoise2(___)`S`

is a matrix with
the same structure as the `S`

output of `wavedec2`

.

`[`

returns the shifts along the row and column dimensions for cycle spinning.
`IMDEN`

,`DENOISEDCFS`

,`ORIGCFS`

,`S`

,`SHIFTS`

] = wdenoise2(___)`SHIFTS`

is
2-by-`(numshifts+1)`

matrix where each
column of ^{2}`SHIFTS`

contains the shifts along the row and column
dimension used in cycle spinning and `numshifts`

is the value of
`CycleSpinning`

.

`[___] = wdenoise2(___,`

returns the denoised image with additional options specified by one or more
`Name,Value`

)`Name,Value`

pair arguments, using any of the preceding
syntaxes.

`wdenoise2(___)`

with no output arguments plots the
original image along with the denoised image in the current figure.

[1] Abramovich, F., Y. Benjamini, D.
L. Donoho, and I. M. Johnstone. “Adapting to Unknown Sparsity by Controlling the False
Discovery Rate.” *Annals of Statistics*, Vol. 34, Number 2, pp.
584–653, 2006.

[2] Antoniadis, A., and G. Oppenheim,
eds. *Wavelets and Statistics*. Lecture Notes in Statistics. New York:
Springer Verlag, 1995.

[3] Donoho, D. L. “Progress in
Wavelet Analysis and WVD: A Ten Minute Tour.” *Progress in Wavelet Analysis
and Applications* (Y. Meyer, and S. Roques, eds.). Gif-sur-Yvette: Editions
Frontières, 1993.

[4] Donoho, D. L., I. M. Johnstone.
“Ideal Spatial Adaptation by Wavelet Shrinkage.”
*Biometrika*, Vol. 81, pp. 425–455, 1994.

[5] Donoho, D. L. “De-noising
by Soft-Thresholding.” *IEEE Transactions on Information Theory*,
Vol. 42, Number 3, pp. 613–627, 1995.

[6] Donoho, D. L., I. M. Johnstone, G.
Kerkyacharian, and D. Picard. “Wavelet Shrinkage: Asymptopia?” *Journal
of the Royal Statistical Society*, *series B*, Vol. 57, No. 2,
pp. 301–369, 1995.

[7] Johnstone, I. M., and B. W.
Silverman. “Needles and Straw in Haystacks: Empirical Bayes Estimates of Possibly
Sparse Sequences.” *Annals of Statistics*, Vol. 32, Number 4, pp.
1594–1649, 2004.