I'm trying to analyze an image, loaded as an array, which currently contains saturated pixels. The analysis function also takes a mask array, which I can easily define to cover the pixels which are saturated or above some threshold value.
What I would like to do is define a masking array which masks both the saturated pixels, and some radius/area around them. I could image looping over every element in the mask array and finding the maximum value within the defined radius for each, but that seems like a very slow way to approach it, and I wondered if there could be anything quicker.
im = np.asarray(Image.open(image))
mask = np.where(im>2000,0,1)
output = some_function(im,mask)