MulticlassLabeler returns different shapes

[jakubczakon]

MulticlassLabeler usually (almost always) returns: (2,300,300) but sometimes you get (1,300,300) which of course causes trouble. It happens because of the mask.max() in the following implementation.

def label_multiclass_image(mask):
    labeled_channels = []
    for label_nr in range(0, mask.max() + 1):
        labeled_channels.append(label(mask == label_nr))
    labeled_image = np.stack(labeled_channels)
    return labeled_image

[taraspiotr]

Yes and this is caused by current categorize_image() implementation. I think it is worth changing, because this kind of inconsistency in shape is not desirable, but it shouldn't cause any error in pipeline running or in the output.