How can i iterate over image pixels in a faster manner in python?

Question

I want to modify a grayscale image in a manner so that I can change the pixel values to black for the top half of the image. I can certainly do this by iterating over in the usual manner like this:

for i in range(0,rows):
  for j in range(0,cols):
    if(condition)
      image[i,j] = 0;

But this is quite slow as I have to do video processing. I can see that I have to use Image.point(), but I am not sure how to implement it. Can someone help me out in this?

Answer 1

This will be much faster if you convert the PIL image to a numpy array first. Here's how you can zero all the pixels with a value below 10:

>>> import numpy as np
>>> arr = np.array(img)
>>> arr[arr < 10] = 0
>>> img.putdata(arr)

Or, as you stated in your comment, here's you'd black out the top half of the image:

>>> arr[:arr.shape[0] / 2,:] = 0

Finally, since you're doing video processing, notice that you don't have to loop over the individual frames either. Let's say you have ten frames of 4x4 images:

>>> arr = np.ones((10,4,4)) # 10 all-white frames
>>> arr[:,:2,:] = 0         # black out the top half of every frame
>>> a
array([[[ 0.,  0.,  0.,  0.],
    [ 0.,  0.,  0.,  0.],
    [ 1.,  1.,  1.,  1.],
    [ 1.,  1.,  1.,  1.]],

   [[ 0.,  0.,  0.,  0.],
    [ 0.,  0.,  0.,  0.],
    [ 1.,  1.,  1.,  1.],
    [ 1.,  1.,  1.,  1.]],
...