GeometricMeanRestoration

GeometricMeanRestoration[img, psf, K,

] restores the degraded image img given psf, the estimate of the degradation process, the scalar K representing the ratio of noise power to signal power, and scalars

, using the geometric mean filter, a generalization of the Wiener filter.

GeometricMeanRestoration[img, psf, S_N,S_F,

] restores the degraded image img given psf, the estimate of the degradation process, S_N and S_F, the squared magnitude Fourier spectra of the noise and undegraded signals, respectively, and scalars

, using the geometric mean filter, a generalization of the Wiener filter.

The GeometricMeanRestoration filter is a generalization of the Wiener filter. With G=DFT[img], H=DFT[psf], and H^* the complex conjugate of H, the result is effectively equal to:

With argument settings

=1, GeometricMeanRestoration returns the same result as WienerRestoration.

The argument img is an array or ImageData expression.

See also User's Guide 8.7.

New in Version 2.

Example

This loads the needed packages.

In[1]:=

This loads an example image and degrades it using horizontal motion blur and additive, normally distributed noise.

In[2]:=

This shows, from left to right, the example image, the degraded image, and restoration results for two different values of the scalars

In[3]:=

Out[3]=

Out[3]=

© 2009 Wolfram Research, Inc. • Terms of Use • Privacy Policy