`scipy.ndimage._measurements:watershed_ift`

Signature

def watershed_ift(input, markers, structure = None, output = None)

Apply watershed from markers using image foresting transform algorithm.

input : array_like: Input.
markers : array_like: Markers are points within each watershed that form the beginning of the process. Negative markers are considered background markers which are processed after the other markers.
structure : structure element, optional: A structuring element defining the connectivity of the object can be provided. If None, an element is generated with a squared connectivity equal to one.
output : ndarray, optional: An output array can optionally be provided. The same shape as input.