In [1]:
%matplotlib inline
import numpy as np
import matplotlib.pyplot as plt
from scipy.misc import imread
In this notebook we explore an application of clustering algorithms to shape segmentation from binary images. We will carry out some exploratory work with a small set of images provided with this notebook. Most of them are not binary images, so we must do some preliminary work to extract he binary shape images and apply the clustering algorithms to them. We will have the opportunity to test the differences between $k$-means and spectral clustering in this problem.
Several images are provided with this notebook:
Select image birds.jpg from file and plot it in grayscale
In [2]:
name = "birds.jpg"
name = "Seeds.jpg"
birds = imread("Images/" + name)
birdsG = np.sum(birds, axis=2)
plt.imshow(birdsG, cmap=plt.get_cmap('gray'))
plt.grid(False)
plt.axis('off')
plt.show()
In [3]:
plt.hist(birdsG.ravel(), bins=256)
plt.show()
Plot the binary image after thresholding.
In [4]:
if name == "birds.jpg":
th = 256
elif name == "Seeds.jpg":
th = 650
birdsBN = birdsG > th
# If there are more white than black pixels, reverse the image
if np.sum(birdsBN) > float(np.prod(birdsBN.shape)/2):
birdsBN = 1-birdsBN
plt.imshow(birdsBN, cmap=plt.get_cmap('gray'))
plt.grid(False)
plt.axis('off')
plt.show()
In [5]:
(h, w) = birdsBN.shape
bW = birdsBN * range(w)
bH = birdsBN * np.array(range(h))[:,np.newaxis]
pSet = [t for t in zip(bW.ravel(), bH.ravel()) if t!=(0,0)]
X = np.array(pSet)
In [6]:
print X
In [7]:
plt.scatter(X[:, 0], X[:, 1], s=5);
plt.axis('equal')
plt.show()
In [8]:
from sklearn.cluster import KMeans
est = KMeans(50) # 4 clusters
est.fit(X)
y_kmeans = est.predict(X)
plt.scatter(X[:, 0], X[:, 1], c=y_kmeans, s=5, cmap='rainbow',
linewidth=0.0)
plt.axis('equal')
plt.show()
In [9]:
from sklearn.metrics.pairwise import rbf_kernel
gamma = 5
sf = 4
Xsub = X[0::sf]
print Xsub.shape
gamma = 0.001
K = rbf_kernel(Xsub, Xsub, gamma=gamma)
In [10]:
plt.imshow(K, cmap='hot')
plt.colorbar()
plt.title('RBF Affinity Matrix for gamma = ' + str(gamma))
plt.grid('off')
plt.show()
In [11]:
from sklearn.cluster import SpectralClustering
spc = SpectralClustering(n_clusters=50, gamma=gamma, affinity='rbf')
y_kmeans = spc.fit_predict(Xsub)
In [12]:
plt.scatter(Xsub[:,0], Xsub[:,1], c=y_kmeans, s=5, cmap='rainbow', linewidth=0.0)
plt.axis('equal')
plt.show()
In [13]:
print X[:,1]
In [ ]: