Revision history [back]

You can use binarization with a threshold on the brightness (L-channel of HSL color space). Then you can apply binarization and opening / closing.

Python sample code

#!/usr/bin/env python

from PIL import Image
import numpy as np
from scipy.misc import imsave
from scipy.ndimage.morphology import binary_closing, binary_opening


def binarize_array(numpy_array, threshold=200):
    """
    Binarize a numpy array.

    Source: https://stackoverflow.com/a/37497975/562769

    Parameters
    ----------
    numpy_array : numpy array
    threshold : int

    Returns
    -------
    numpy_array
    """
    for i in range(len(numpy_array)):
        for j in range(len(numpy_array[0])):
            if numpy_array[i][j] > threshold:
                numpy_array[i][j] = 255
            else:
                numpy_array[i][j] = 0
    return numpy_array

im = Image.open("14810202126365282.jpg")
im = im.convert('L')
im = np.array(im)
im = binarize_array(im, 200)
im = binary_closing(im, structure=np.ones((10, 10)))
im = binary_opening(im, structure=np.ones((10, 10)))
imsave("out.png", im)

Sample mask for the book