Separating objects that are collated

Hi there,

I’m struggling to separate the following cells with no luck. I need the Roi of each cell. I have also nuclei mask in case it helps.
A - 01(fld 003 wv FITC - FITC)CellsOutlines.tiff (4.0 MB)
A - 01(fld 003 wv FITC - FITC)NucsOutlines.tiff (4.0 MB)

Any ideas?
Thank you

It would help the forum if you could post the original image and not the outlines.

However search this forum for e.g., a seperation of touching objects, e.g.,:

A popular method is Watershed, e.g.,:


I didn’t include the original image because the idea is to post-process the result of cellprofiler segmentation. Although the cells are well segmented, whenever two contours are touching the analyse particles join them in the same ROI (see below). I need a Roi for each cell.

Clipboard-1.tif (2.1 MB)

Thank you

Once again the original image would be helpful. Since you did it in cellprofiler I added the cellprofiler tag so that people which use the software might help you to get a better result.

Please note that it is important to know which analysis chain you used to help effectively.

A reproducible example speeds up an effective help.

1 Like

I just would like sharing the solution that I came out after some trials.

//identify correct contours based on cell nucleus
	run("Fill Holes");
	run("Find Maxima...", "prominence=1 strict exclude light output=[Single Points]");
	run("Fill Holes");
	run("Marker-controlled Watershed", "input=" + cells +" marker=marker mask=None binary calculate use");
	setAutoThreshold("Default dark");
	setThreshold(0, 0);
	run("Convert to Mask");
	imageCalculator("Subtract create", cells,"watershed_lines");

It may help someone!