BlockLee

BlockLee[img, {n₁, n₂}, var] is a nonlinear noise reduction operator.

BlockLee returns the best mean squared error estimate of each image pixel based on the statistics of its centered neighborhood of dimensions n₁×n₂. Effectively, this is a pixel-wise adaptive Wiener method.

This shows the geometry of a 3 × 3 neighborhood centered on f(r,c):

The argument var is a user-supplied estimate of the variance of the additive noise signal.

The argument img may be a vector, matrix, or ImageData expression.

See also User's Guide 5.5.

Modified in Version 2.

Example

This loads the package.

In[1]:=

This generates a data array with uniformly distributed additive noise.

In[2]:=

Here is the noisy data.

In[3]:=

Out[3]=

Out[3]=

This demonstrates noise suppression using BlockLee with 3 × 3 neighborhoods.

In[4]:=

In[5]:=

Out[5]=

Out[5]=

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