skimage.filters.rank.mean and scipy.signal.convolve2d

rank_mean.png

rank_mean.py

import matplotlib.pyplot as plt
import numpy as np
from scipy.signal import convolve2d
from skimage import data
from skimage.filters import rank

image = data.coins()
K = np.ones((11, 11))

rank_mean = rank.mean(image, selem=K)
naive_convolve = np.array(convolve2d(image, K, mode="same"), dtype=float) / float(K.sum())

plots = [
 ("rank.mean", rank_mean, (0, 255), plt.cm.gray),
 ("convolve2d", naive_convolve, (0, 255), plt.cm.gray),
 ("Input Image", image, (0, 255), plt.cm.gray),
 ("abs(convolve2d - rank.mean)", abs(naive_convolve - rank_mean), (0, 1), plt.cm.jet)
]

fig, axes = plt.subplots(nrows=2, ncols=2, figsize=(8, 8),
                         sharex=True, sharey=True)

for ax, (title, img, (vmin, vmax), cmap) in zip(axes.ravel(), plots):
    im = ax.imshow(img, cmap=cmap, vmin=vmin, vmax=vmax)
    ax.set_title(title)
    ax.set_adjustable('box-forced')
    ax.axis('off')

plt.tight_layout()
plt.savefig("rank_mean.png")
plt.show()