CODEWITHSUNDEEP
pythonimagergb
NoobCat
NoobCat

Reputation: 91

How to invert the green channel of an image in Python?

How do I invert one channel of an RGB image? In this case my image is a normal map for a 3D engine, saved in an image format like JPEG or TIFF, and I want to invert the green channel — that is, completely reverse the highs and lows in the green band.

Upvotes: 1

Views: 2309

Answers (3)

CrazyChucky
CrazyChucky

Reputation: 3518

You can accomplish this by installing and using Pillow, which can work with most image formats (JPEG, PNG, TIFF, etc.).

from PIL import Image
from PIL.ImageChops import invert

image = Image.open('test.tif')
red, green, blue = image.split()
image_with_inverted_green = Image.merge('RGB', (red, invert(green), blue))
image_with_inverted_green.save('test_inverted_green.tif')

After loading the image from your file, split into its channels with Image.split, invert the green channel/image with ImageChops.invert, and then join it together with the original red and blue bands into a new image with Image.merge.

If using a format that's encoded in something other than RGB (such as PNG, which has an additional transparency channel), the image-opening line can be amended to:

image = Image.open('test.png').convert('RGB')

Testing with this image:

test.tif

Produces this:

test_with_inverted_green.tif

(ImageChops, by the way, looks like a very odd term, but it's short for "image channel operations".)

Upvotes: 4

Fahim Aslam
Fahim Aslam

Reputation: 1

T(r) = L – 1 – r

L=256

L-1=255(Max)

r=Each pixel of the image

s=255-r

s= T(r) =255-r

Upvotes: -3

modesitt
modesitt

Reputation: 7210

I assume you have a numpy (or torch tensor) image - you can index on the green channel (assuming channels are your last dimension)

img[:, :, 1] = 255 - img[:, :, 1]

I am assuming by invert you want 0 -> 255 and 255 -> 0

Upvotes: 0

Related Questions