ImageJ Documentation Wiki

Filter Lab stack image using a selective gaussian blur. It compute a ponderated mean of in the neighborhood pixel. For each neighborhood pixel 'q' of computed pixel 'p', the associated weight 'q' is computed with a gaussian function and the luminance difference between luminance 'p' and luminance 'q'. If difference is zero, weight is 1. If difference is bigger weight tend towards zero. Only chrominance slices are filtered ( slices a and b ).

1. To use this filter, you must first (step 1) convert your RGB image to Lab space color with this plugin : Converting RGB to Lab
2. Then, use this filter (step 2)
3. Finally, convert back the Lab space color image to RGB (step 3) with this plugin : Converting Lab to RGB

NB:

• This plugin works well after this other one : Line Lab selective median filtering

lab_nonlocal_mean_with_same_pixel_luminance_neighborhood_v1_01.jar

It use no additionnal libraries.

Copy the .jar file to your plugins directory and restart “ImageJ”.

You can use the script by menu “ImageJ / Plugins / Image/ Color / Lab non-local mean with same pixel luminance neighborhood (v1.00)”.

lab_nonlocal_mean_with_same_pixel_luminance_neighborhood_v1_01.zip

Vincent Vansuyt.

Contact me for any questions or suggestions : vincent dot vansuyt at edf dot fr (preferred language : french)

GPL distribution (see licence ).

Version v1.00 : first version

Version v1.01 :

• adding timing
• code improvement (40% faster)

None

• Conversions between RGB color space and Lab color space
• Line Lab selective median filtering
• Find minimum and maximum
• Set minimum and maximum values for LUT