Upper bound on threshold


I have tried a few times to limit the upper bound of thresholding when identifying primary objects in tissue slices. This is convenient to eliminate artefacts, as they tend to be saturated. Somehow,this does not seem to change anything to the segmentation. I have tried lower and upper bounds of 0.19 and 0.191 and can segment all the objects up to 1. The lower bound definitely works the upper one does not really do anything. Did I misunderstand something?
Thank you,

Hi Marc,

What I think you’re asking is whether the upper bound can serve to eliminate bright objects by removing objects brighter than a given upper bound, correct? If I’m mistaken, please correct me.

If I’m correct, the upper bound actually does something different. It prevents the automatic threshold from being calculated as higher than your given limit. So if you have an object that is greater in intensity than the upper bound, it will still get thresholded and segmented, just that the threshold is not as high as it would have been otherwise.

If you are looking to actually eliminate bright objects that were detected, you can use FilterObjects to remove them on the basis of intesnity. Alternately, you can use ApplyThreshold to mask out pixels higher than a given threshold; in this case, the output is an image (as opposed to objects) which you can then use as input to IdentifyPrimaryObjects.


Hi Mark,

perfect, you answer my question fully. I’ll try it.

Thank you,