opencv - 如何从一维子数组的数组中选择具有最小和最大数量的白色像素的子数组？

Question

如何从一维 numpy 子数组的数组中选择子数组？

def distancesplit(self):
  img = np.asarray(Image.open("testtwo.tif").convert('L'))                            
  img = 1 * (img < 127)
  area = (img == 0).sum()
  areasplit = np.split(img, 24) # splitting an image array into 24
  print areasplit
for i in areasplit: # compute white pixel area in splitted parts-arrays
  result = (i == 0).sum()
  print result
for i in areasplit: #converting 2D array inro 1D 
  b = i.ravel()
# Now I need o select from "b" 1D array the subarrays  with minimal and maximal number of white pixels. I need it for distance metrics between these subarrays. Thank you.

Answer 1

在每个子矩阵（在这种情况下为行或列）上使用 sum 函数（http://docs.opencv.org/modules/core/doc/operations_on_arrays.html#sum），然后从结果中找到最小值和最大值。