mirror of
https://github.com/wassname/scikit-image.git
synced 2026-06-28 03:03:07 +08:00
Vighnesh Birodkar
91 lines
2.8 KiB
Python
91 lines
2.8 KiB
Python
"""
|
|
==============
|
|
Blob Detection
|
|
==============
|
|
|
|
Blobs are bright on dark or dark on bright regions in an image. In
|
|
this example, blobs are detected using 3 algorithms. The image used
|
|
in this case is the Hubble eXtreme Deep Field. Each bright dot in the
|
|
image is a star or a galaxy, so we are literally counting stars.
|
|
|
|
Laplacian of Gaussian (LoG)
|
|
-----------------------------
|
|
This is the most accurate and slowest approach. It computes the Laplacian
|
|
of Gaussian images with successively increasing standard deviation and
|
|
stacks them up in a cube. Blobs are local maximas in this cube. Detecting
|
|
larger blobs is especially slower because of larger kernel sizes during
|
|
convolution. Only bright blobs on dark backgrounds are detected.
|
|
|
|
Difference of Gaussian (LoG)
|
|
----------------------------
|
|
This is a faster approximation of LoG approach. In this case the image is
|
|
blurred with increasing standard deviations and the difference between
|
|
two successively blurred images are stacked up in a cube. This method
|
|
suffers from the same disadvantage as LoG approach for detecting larger
|
|
blobs. Blobs are again assumed to be bright on dark.
|
|
|
|
Determinant of Hessian (DoH)
|
|
----------------------------
|
|
This is the fastest approach. It detects blobs by finding maximas in the
|
|
matrix of the Determinant of Hessian of the image. The detection speed is
|
|
independent of the size of blobs as internally the implementation uses
|
|
box filters instead of convolutions. Bright on dark as well as dark on
|
|
bright blobs are detected.The downside is that small blobs (<3px) are not
|
|
detected accurately.
|
|
|
|
"""
|
|
|
|
from matplotlib import pyplot as plt
|
|
from skimage import data
|
|
from skimage.feature import blob_dog, blob_log, blob_doh
|
|
from math import sqrt
|
|
from skimage.color import rgb2gray
|
|
|
|
image = data.hubble_deep_field()[0:500, 0:500]
|
|
image_gray = rgb2gray(image)
|
|
|
|
blobs = blob_log(image_gray, max_sigma=30, num_sigma=10, threshold=.1)
|
|
|
|
fig = plt.figure()
|
|
ax = fig.add_subplot(1, 1, 1)
|
|
ax.set_title('Laplacian of Gaussian')
|
|
plt.imshow(image)
|
|
|
|
for blob in blobs:
|
|
y, x = blob[0], blob[1]
|
|
r = blob[2] * sqrt(2)
|
|
c = plt.Circle((x, y), r, color='#00ff00', lw=2, fill=False)
|
|
ax.add_patch(c)
|
|
|
|
|
|
blobs = blob_dog(image_gray, max_sigma=30, threshold=.1)
|
|
|
|
fig = plt.figure()
|
|
ax = fig.add_subplot(1, 1, 1)
|
|
ax.set_title('Difference of Gaussian')
|
|
plt.imshow(image)
|
|
|
|
for blob in blobs:
|
|
y, x = blob[0], blob[1]
|
|
r = blob[2] * sqrt(2)
|
|
c = plt.Circle((x, y), r, color='#ffff00', lw=2, fill=False)
|
|
ax.add_patch(c)
|
|
|
|
|
|
blobs = blob_doh(image_gray, max_sigma=30, threshold=.01)
|
|
|
|
fig = plt.figure()
|
|
ax = fig.add_subplot(1, 1, 1)
|
|
ax.set_title('Determinant of Hessian')
|
|
plt.imshow(image)
|
|
|
|
|
|
for blob in blobs:
|
|
y, x = blob[0], blob[1]
|
|
r = blob[2]
|
|
c = plt.Circle((x, y), r, color='#ff0000', lw=2, fill=False)
|
|
ax.add_patch(c)
|
|
|
|
|
|
plt.show()
|