Point to Watershed?

[Nathaniel Ryckman]

I was wondering what the function of Watershed was compared to the Maximum Finder?

What is the point in converting a gray scale image to a Euclidean Distance Map before flooding? Is it to try to compensate for over segmentation from the Maximum Finder?

Thanks for the clarification.

[Nathaniel Ryckman]

After looking through the MaximumFinder class, I think the Watershed program became useful because the MaximumFinder class oversegments cells due to its heavy reliance on local maxima and the way it binarizes the image (spreads 255 pixels).

In other words, by using thresholding->binarizing->watershed on an image, much of the images complexity is removed/destroyed. It simplifies the problem, but, consequently, it also doesn't represent the problem accurately.