minmax

Input: a matrix containing a gray-scale image to be filtered by min/max algorithm.
NSteps: number of steps to perform. Default is 10.
StepSize: the step increment for the iterative procedure. Default is 0.05.
Adapt: indicates if the algorithm should adapt itself to the local image gray level or if it considers the NonAdaptThreshold value for defining light and dark regions. Default is FALSE.
NonAdaptThreshold: If Adapt is FALSE, intensity values greater than NonAdaptThreshold will be considered as light regions.
IntMaskSize: Size of the Internal window in which curvature values will be taken into account for deciding Min or Max curvature flow. Default is 1.
ExtMaskSize: Size of the External window in which curvature values will be taken into account for deciding Min or Max curvature flow in the Adapt mode. Default is 0.
Output: a matrix containing the filtered image.

Function minmax filters a gray-scale image using curvature-guided surface evolution. Object borders remain sharp while low-scale noise is removed.

M = gray_imread(SIPDIR+'images/noisypoly.bmp');
subplot(1,2,1);
imshow(M);
new_M = minmax(M, NSteps=30);
subplot(1,2,2);
imshow(new_M);

The latest version of the Scilab Image Processing toolbox can be found at

http://siptoolbox.sourceforge.net