Limeseg results extraction


I have been playing around with the Plugin LimeSeg and obtained very convincing results for segmentation. However after reading the wiki and browsing through the forum I am still confused on how to extract results of the segmentation and export them.

Thanks for your help!


Woops, just saw your message, sorry! Just ping me as you did in the other thread, and I’ll show up faster. Did you manage to get the results out ? There are 3 ways to get outputs:

  • The Commands found in the plugins (Sphere Seg, Skeleton Seg) will output basic information like surface and volume per object. This is displayed in a standard ImageJ results table.
  • You can save a folder containing ply files where each file contains a point cloud and a mesh if it has been computed. This can be achieved with the LimeSeg GUI as described here ( The GUI is a bit annoying. You need to set an empty destination folder AND THEN click on the save button.
  • You can script within FIJI in groovy, and basically access all properties of the objects. For instance this script example ( crops all points above a certain x coordinate.

It loops through all Cells, timepoint 1, and all points.

Here’s for instance a groovy script which writes the coordinates of all points for all objects of the timepoint 1, as well as their corresponding unit normal vector:

import eu.kiaru.limeseg.LimeSeg;
// Objects used by LimeSeg
import eu.kiaru.limeseg.struct.Cell;
import eu.kiaru.limeseg.struct.CellT;
import eu.kiaru.limeseg.struct.DotN;
import eu.kiaru.limeseg.struct.Vector3D;

for (Cell c:LimeSeg.allCells) {
	CellT ct = c.getCellTAt(1)
	for (DotN dn:ct.dots) {
		print("P=\t"+dn.pos.x+"\t"+dn.pos.y+"\t"+dn.pos.z+"\t N=\t"+dn.Norm.x+"\t"+dn.Norm.y+"\t"+dn.Norm.z+"\n");

If you want to be notified without people having to mention you, you can also subscribe to the #limeseg tag (click the circle icon in the top right on the tag feed page).

Me and other moderators on the forum (are trying to) tag new posts if the original posters tagged them insufficiently.