Python equivalent to Matlab funciton 'imfill' for grayscale?

Question

Is there an implementation using OpenCV or scikit-image that is equivalent to Matlab's grayscale image imfill funciton (i.e. grayscale hole filling)?

See the imfill section for grayscale (I2= imfill(I)) in the following example link matlab_imfill. Or see image: matlab_tire_ex

Here's a link to the tire image in the example

I've been trying to replicate the Matlab output using scipy.ndimage.grey_closing function with varying the size parameter, but have not been successful.

I'm using Python 3.5.

Answer 1

Matlab infill() in turn uses a function IM = imreconstruct(marker,mask)

Scikit-image has a similar function... skimage.morphology.reconstruction(seed, mask, method='dilation', selem=None, offset=None)

The algorithm is detailed in Soille, P., Morphological Image Analysis: Principles and Applications, Springer-Verlag, 1999, pp. 208-209. section 6.3.7 section "Fillhole"

import numpy as np
from skimage.morphology import reconstruction
import matplotlib.pyplot as plt
from skimage.io import imread, imsave


# Use the matlab reference Soille, P., Morphological Image Analysis: Principles and Applications, Springer-Verlag, 1999, pp. 208-209.
#  6.3.7  Fillhole
# The holes of a binary image correspond to the set of its regional minima which
# are  not  connected  to  the image  border.  This  definition  holds  for  grey scale
# images.  Hence,  filling  the holes of a  grey scale image comes down  to remove
# all  minima  which  are  not  connected  to  the  image  border, or,  equivalently,
# impose  the  set  of minima  which  are  connected  to  the  image  border.  The
# marker image 1m  used  in  the morphological reconstruction by erosion is set
# to the maximum image value except along its border where the values of the
# original image are kept:

img = imread("tyre.jpg")

seed = np.ones_like(img)*255
img[ : ,0] = 0
img[ : ,-1] = 0
img[ 0 ,:] = 0
img[ -1 ,:] = 0
seed[ : ,0] = 0
seed[ : ,-1] = 0
seed[ 0 ,:] = 0
seed[ -1 ,:] = 0


fill = reconstruction(seed, img, method='erosion')

f, (ax0, ax1) = plt.subplots(1, 2,
    subplot_kw={'xticks': [], 'yticks': []},
    figsize=(12, 8))
ax0.imshow(img)
ax1.imshow(fill)
plt.show()

Link to tyre image and filled image

Answer 2

Two versions of the flood-fill algorithm have been implemented in Python here:

http://arcgisandpython.blogspot.de/2012/01/python-flood-fill-algorithm.html

The first, simpler one contained two undefined variables, but here is a working version:

import numpy as np
import scipy as sp
import scipy.ndimage

def flood_fill(test_array,h_max=255):
    input_array = np.copy(test_array) 
    el = sp.ndimage.generate_binary_structure(2,2).astype(np.int)
    inside_mask = sp.ndimage.binary_erosion(~np.isnan(input_array), structure=el)
    output_array = np.copy(input_array)
    output_array[inside_mask]=h_max
    output_old_array = np.copy(input_array)
    output_old_array.fill(0)   
    el = sp.ndimage.generate_binary_structure(2,1).astype(np.int)
    while not np.array_equal(output_old_array, output_array):
        output_old_array = np.copy(output_array)
        output_array = np.maximum(input_array,sp.ndimage.grey_erosion(output_array, size=(3,3), footprint=el))
    return output_array