051-Kmeans_using_opencv.py

#!/usr/bin/env python
__author__ = "Sreenivas Bhattiprolu"
__license__ = "Feel free to copy, I appreciate if you acknowledge Python for Microscopists"

# https://www.youtube.com/watch?v=6CqRnx6Ic48



import numpy as np
import cv2

img = cv2.imread("images/BSE_Image.jpg")


# Convert MxNx3 image into Kx3 where K=MxN
img2 = img.reshape((-1,3))  #-1 reshape means, in this case MxN

#We convert the unit8 values to float as it is a requirement of the k-means method of OpenCV
img2 = np.float32(img2)

#Define criteria, number of clusters and apply k-means
#When this criterion is satisfied, the algorithm iteration stops. 
#cv.TERM_CRITERIA_EPS — stop the algorithm iteration if specified accuracy, epsilon, is reached.
#cv.TERM_CRITERIA_MAX_ITER — stop the algorithm after the specified number of iterations, max_iter.
#cv.TERM_CRITERIA_EPS + cv.TERM_CRITERIA_MAX_ITER — stop the iteration when any of the above condition is met.
#Max iterations, in this example 10. 
#Epsilon, required accuracy, in this example 1.0

criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 10, 1.0)

# Number of clusters
k = 4

# Number of attempts, number of times algorithm is executed using different initial labelings.
#Algorithm return labels that yield best compactness.
#compactness : It is the sum of squared distance from each point to their corresponding centers.

attempts = 10

#other flags needed as inputs for K-means
#Specify how initial seeds are taken.
#Two options, cv.KMEANS_PP_CENTERS and cv.KMEANS_RANDOM_CENTERS

ret,label,center=cv2.kmeans(img2, k, None, criteria, attempts, cv2.KMEANS_PP_CENTERS)

#cv2.kmeans outputs 2 parameters.
#1 COmpactness. 
#2 Labels: Label array.
#3 Center. the array of centers of clusters. For k=4 we will have 4 centers.
#For RGB image, we will have center for each image, so tota 4x3 = 12.
#Now convert center values from float32 back into uint8.
center = np.uint8(center) 

#Next, we have to access the labels to regenerate the clustered image
res = center[label.flatten()]
res2 = res.reshape((img.shape)) #Reshape labels to the size of original image
cv2.imwrite("images/segmented.jpg", res2)


"""
#Now let us visualize the output result
figure_size = 15
plt.figure(figsize=(figure_size,figure_size))
plt.subplot(1,2,1),plt.imshow(img2)
plt.title('Original Image'), plt.xticks([]), plt.yticks([])
plt.subplot(1,2,2),plt.imshow(res2)
plt.title('Segmented Image when K = %i' % k), plt.xticks([]), plt.yticks([])
plt.show()
"""