norm_to_uniform#

Import packages#

import numpy as np
import porespy as ps
import scipy.ndimage as spim
import matplotlib.pyplot as plt
import skimage
ps.visualization.set_mpl_style()

Generate image for testing#

im = np.random.rand(200, 200)
strel = ps.tools.ps_disk(20, smooth=False)
im = spim.convolve(im, weights=strel)
fig, ax = plt.subplots(1, 2, figsize=[8, 4])
ax[0].axis(False)
ax[0].imshow(im)
ax[1].hist(im.flatten(), edgecolor='k', bins=25)
ax[1].set_xlabel('Value')
ax[1].set_ylabel('Counts');
../../../_images/c3b647a1e293ed987b5795fd885490b97ff083cefa6103178684a9c843789540.png

Demonstrate function#

The correlated noise field generated above has approximatetly normally distributed values. It’s not perfectly normal, but it’s pretty close. This can be converted to uniformly distributed values as follows:

im1 = ps.tools.norm_to_uniform(im=im)
fig, ax = plt.subplots(1, 2, figsize=[8, 4])
ax[0].axis(False)
ax[0].imshow(im1)
ax[1].hist(im1.flatten(), edgecolor='k', bins=25)
ax[1].set_xlabel('Value')
ax[1].set_ylabel('Counts');
../../../_images/01849e26bada5d1b1088bdaf98648b04339b0d39c5784dd0853b2a5c93dccde3.png

scale#

The output can be scale to a specific range:

im2 = ps.tools.norm_to_uniform(im=im, scale=[0, 1])
fig, ax = plt.subplots(1, 2, figsize=[8, 4])
ax[0].axis(False)
ax[0].imshow(im2)
ax[1].hist(im2.flatten(), edgecolor='k', bins=25)
ax[1].set_xlabel('Value')
ax[1].set_ylabel('Counts');
../../../_images/0fe4584ebd8c40bdceb19db204080a1fbfae4105668e42a17cf117b0d081da22.png