trim_saddle_points#

trim_saddle_points function is a filter which removes peaks that were mistakenly identified because they lie on a saddle or a ridge in the distance transform.

import matplotlib.pyplot as plt
from edt import edt

import porespy as ps

ps.visualization.set_mpl_style()

dt#

The distance transform of a test image is detetmined using the edt method.

im = ps.generators.blobs(shape=[200, 200], blobiness=[2, 0.5], porosity=0.65)
dt = edt(im)

plt.figure(figsize=[6, 6])
plt.axis(False)
plt.imshow(dt / im);
../../../_images/9d425530686ade427c8aec0409c0b4c4e38b00c757291eaa58d4e9459e7f21f8.png

peaks#

peaks is found using the find_peaks filter on the distance transform. Peaks that lie on a saddle or a ridge are removed.

peaks = ps.filters.find_peaks(dt, r_max=4)

x = ps.filters.trim_saddle_points(peaks=peaks, dt=dt)

fix, ax = plt.subplots(1, 2, figsize=[12, 12])
ax[0].axis(False)
ax[0].imshow(peaks / im)
ax[0].set_title("Before", fontdict={"fontsize": 18})
ax[1].axis(False)
ax[1].imshow(x / im)
ax[1].set_title("After", fontdict={"fontsize": 18});
../../../_images/b0cf214a764d915860002d3377510ea25c6130a66aea17412a514110a8d07e88.png