OOF2: The Manual

OOF.Image.AutoGroup
	6.2. Menus

Name

OOF.Image.AutoGroup — Create a pixel group for each color in the image.

Synopsis

OOF.Image.AutoGroup(image, name_template)

Details

Parent Menu: OOF.Image
Callback: function createPixelGroups in module ooflib.image.IO.imagemenu
Parameters:

image

Type: The path to an existing Image object.

name_template

Name for the pixel groups. %n is replaced by a number, %c by the pixel color. Type: A character string.

Description

If a Microstructure is being built from an Image which has a one-to-one correspondence between pixel color and Material, then the most convenient way to assign Materials to pixels is to automatically create a PixelGroup for each color, and then to assign Materials to the pixels in each pixel group. The AutoGroup command performs the first part of this process, creating pixel groups for each color of pixel in the given Image.

The name_template parameter specifies how the names for the automatically created groups will be generated. It should be a string containing either %c or %n or both. When naming a group, %c will be replaced by the color of the pixels in the group, and %n will be replaced by an integer. The colors are written in the slightly uncomfortable but compact form #RRGGBB, where RR, GG, and BB are the hexadecimal values of the 8-bit RGB color channels. It may be advisable to manually rename the groups.

This command operates on all pixels in the Image, not just the active ones.

	Caution
This command does not work well on noisy experimental micrographs, because pixels whose colors differ only due to noise will be put into different groups. It also doesn't work well on artifical micrographs produced by drawing programs, if the programs use antialiasing to smooth lines and boundaries. Even if an antialiased image appears to contain only a few colors, it will generally have many slightly different colors at the borders between regions, and AutoGroup will create a large number of small groups. Better ways of creating pixel groups from noisy or antialiased images are Select pixels with a tool such as OOF.PixelSelection.Color_Range and assign them to a group manually with OOF.PixelGroup.AddSelection. Use OOF.PixelGroup.AutoGroup (the Auto button on the Microstructure page), which groups similarly colored pixels without requiring that they be identical. Create a clean image (i.e, one with exactly one color per material) using the segmenting tools in other image processing packages, load the clean image into OOF2, and then use AutoGroup on it.

Caution

This command does not work well on noisy experimental micrographs, because pixels whose colors differ only due to noise will be put into different groups. It also doesn't work well on artifical micrographs produced by drawing programs, if the programs use antialiasing to smooth lines and boundaries. Even if an antialiased image appears to contain only a few colors, it will generally have many slightly different colors at the borders between regions, and AutoGroup will create a large number of small groups.

Better ways of creating pixel groups from noisy or antialiased images are

Select pixels with a tool such as OOF.PixelSelection.Color_Range and assign them to a group manually with OOF.PixelGroup.AddSelection.
Use OOF.PixelGroup.AutoGroup (the Auto button on the Microstructure page), which groups similarly colored pixels without requiring that they be identical.
Create a clean image (i.e, one with exactly one color per material) using the segmenting tools in other image processing packages, load the clean image into OOF2, and then use AutoGroup on it.


OOF.Image.Modify.ThresholdImage		OOF.Image.Copy

OOF2: The Manual

Name

Synopsis

Details

Description

See Also