How to stack multiple numpy 2d arrays into one 3d array?

Question

Here's my code:

img = imread("lena.jpg")
for channel in range(3):
    res = filter(img[:,:,channel], filter)
    # todo: stack to 3d here

As you can see, I'm applying some filter for every channel in the picture. How do I stack them back to a 3d array? (= the original image shape)

Thanks

Collect the res in a list. np.stack(alist, axis=2) can be used to join them into an array on a last channel axis. — hpaulj
– hpaulj, Commented Dec 21, 2018 at 19:14
You have filter both as a function and variable, beware of that. Also what kind of filter are you using? I don't see your filter variable is changing by the channel, so why not do a 3D filter and avoid the loop? — anishtain4
– anishtain4, Commented Dec 21, 2018 at 19:31

Dani Mesejo · Accepted Answer · 2018-12-21 19:19:10Z

1

You could use np.dstack:

import numpy as np

image = np.random.randint(100, size=(100, 100, 3))

r, g, b = image[:, :, 0], image[:, :, 1], image[:, :, 2]

result = np.dstack((r, g, b))

print("image shape", image.shape)
print("result shape", result.shape)

Output

image shape (100, 100, 3)
result shape (100, 100, 3)

answered Dec 21, 2018 at 19:19

Dani Mesejo

62.2k6 gold badges56 silver badges86 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

SpghttCd · Accepted Answer · 2018-12-21 19:49:08Z

1

I'd initialize a variable with the needed shape before:

img = imread("lena.jpg")
res = np.zeros_like(img)     # or simply np.copy(img)
for channel in range(3):
    res[:, :, channel] = filter(img[:,:,channel], filter)

edited Dec 21, 2018 at 19:49

answered Dec 21, 2018 at 19:41

SpghttCd

10.9k2 gold badges23 silver badges28 bronze badges

Collectives™ on Stack Overflow

How to stack multiple numpy 2d arrays into one 3d array?

2 Answers 2

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related