How to resize an image and keep backround empty using OpenCV

Question

I have the following 256x256 image:

I want to resize it to 100x100 pixels using OpenCV:

img = cv2.imread('image.png')

p = 100/256
new_width = int( img.shape[1] * p ) 
new_height = int( img.shape[0] * p ) 

resized = cv2.resize(img, (new_width, new_height))

What I get after execution of the code above:

As you can see, its background colour changed to black. But I want to keep the backgound empty. What can I do?

you need to load the image with 4 channels. 3 for the colors 1 for transparency/alpha. im = cv2.imread('image.png', cv2.IMREAD_UNCHANGED) should work. if you allready have data: im = cv2.cvtColor(rgb_data, cv2.COLOR_RGB2RGBA) — Bindestrich
– Bindestrich, Commented Jul 15, 2024 at 11:12
you WILL get artefacts at edges between transparent and opaque areas. if you're lucky, those edges will just appear darker, but they may be tinted with wild colors. that depends on what colors the transparent pixels happen to be given. — Christoph Rackwitz
– Christoph Rackwitz, Commented Jul 15, 2024 at 12:18
This question is similar to: Set white background for a png instead of transparency with OpenCV. If you believe it’s different, please edit the question, make it clear how it’s different and/or how the answers on that question are not helpful for your problem. — Christoph Rackwitz
– Christoph Rackwitz, Commented Jul 15, 2024 at 12:21

JayPeerachai · Accepted Answer · 2024-07-15 11:12:31Z

1

Try read the image with cv2.IMREAD_UNCHANGED:

img = cv2.imread('image.png', cv2.IMREAD_UNCHANGED)

Reading with cv2.IMREAD_UNCHANGED to ensure the alpha channel (transparency) is preserved.

answered Jul 15, 2024 at 11:12

JayPeerachai

3,8523 gold badges19 silver badges34 bronze badges

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

1 Comment

Christoph Rackwitz Over a year ago

that's still not going to solve the artefacts that'll happen due to the resize()

Collectives™ on Stack Overflow

How to resize an image and keep backround empty using OpenCV

1 Answer 1

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related