Would it be possible to modify the “discard objects on the border of the image” option so that the top & left perimeter edge object measurements are retained? This would ensure that cells split by a camera FOV edge are not double counted when images are acquired on a FL tissue slide scanner? Is this a feasible option?
I’m pretty sure that such a change wouldn’t be able to make it into the upcoming release if we were to implement it.
But even so, while I understand the problem (and have come across it myself), I’m not sure how that would be best implemented. For example, if you had an image divided into quadrants, you would want the user to specify which edges do/don’t get counted depending on the quadrant. Then the problem becomes specifying how to provide a site ID to IdentifyPrimary such that it does the right thing for a different site (and possibly a different set of edges to ignore).
(Not saying it’s impossible, mind you, just tricky)