OpenCV kmeans 错误：(-215) python

Question

Error: error: (-215) (unsigned)labels[i] < (unsigned)K in function kmeans

self.cluster[i] represents some calculated pixel position.

img = numpy.asarray(img)
Z = img.reshape((-1,3))
Z = numpy.float32(Z)
criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 10, 1.0)
#prepare centers
labels = numpy.zeros(k)
for i in range(k):
    labels[i] = self.cluster[i].j * img.shape[0] + self.cluster[i].i
print(Z)
ret, label, center = cv2.kmeans(Z, k, None, criteria, 10, cv2.KMEANS_USE_INITIAL_LABELS, labels)

I have read the other posts about the same error in c++, but I could not get it to work. Please help!

EDIT1:

import numpy as np
import sys
sys.path.append('/usr/local/lib/python3.5/site-packages')
import cv2

img = cv2.imread('image.jpg')
img = cv2.resize(img, (90,90))
Z = img.reshape((-1,3))

# convert to np.float32
Z = np.float32(Z)

# define criteria, number of clusters(K) and apply kmeans()
criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 10, 1.0)
K = 4

labels = np.zeros(K)
labels[0] = 1
labels[1] = 100
labels[2] = 500
labels[3] = 1000
print(labels)
print(Z.shape)
ret,label,center=cv2.kmeans(Z,K,None,criteria,10,cv2.KMEANS_USE_INITIAL_LABELS, labels)

# Now convert back into uint8, and make original image
center = np.uint8(center)
res = center[label.flatten()]
res2 = res.reshape((img.shape))

cv2.imshow('res2',res2)
cv2.waitKey(0)
cv2.destroyAllWindows()

You can run this code alone to check what's wrong. By resizing the image to less than 90x90 causes it to crash. Anything bigger than 91x91 works. Can anyone explain why and maybe how to fix it?

Answer 1

使用cv2.KMEANS_PP_CENTERS或cv2.KMEANS_RANDOM_CENTERS而不是cv2.KMEANS_USE_INITIAL_LABELS时，我的问题解决了。