CODEWITHSUNDEEP
pythonpython-imaging-library
jsmith
jsmith

Reputation: 295

Use Python Imaging Library to isolate a single channel

So say I have an image, and I want to make it so that only the red channel shows up, and the image looks red, how would I do this using PIL? Thanks.

Upvotes: 9

Views: 18665

Answers (4)

pallgeuer
pallgeuer

Reputation: 1366

For people who are looking to extract a single channel from an image (as opposed to generating an image with R, G and B channels, but with the G and B channels all zero), you can do:

img = Image.open("image.jpg")
red = img.getchannel('R')
# OR
red = img.getchannel(0)  # Avoids internal lookup of 'R' in img.getbands()

Upvotes: 1

seequ
seequ

Reputation: 456

Late here, but because we're dealing with numpy, we can use boolean indexing.

def channel(img, n):
    """Isolate the nth channel from the image.

       n = 0: red, 1: green, 2: blue
    """
    a = np.array(img)
    a[:,:,(n!=0, n!=1, n!=2)] *= 0
    return Image.fromarray(a)

Upvotes: 2

Anoyz
Anoyz

Reputation: 7691

You can use the Image.split() operation from PIL to separate the image into bands:

img = Image.open("image.jpg")
red, green, blue = img.split()

If the image has an alpha channel (RGBA) the split function will aditionaly return that. More information here.

Upvotes: 16

jsmith
jsmith

Reputation: 295

I found the answer. Instead of using im.split(), which converted the band to grayscale, I should've converted the Image to an array,multiplied the bands I don't want by 0, and then turned it back to an Image object.

Importing Image and numpy, I did the following:

a = Image.open("image.jpg")
a = numpy.array(a)
a[:,:,0] *=0
a[:,:,1] *=0
a = Image.fromarray(a)
a.show()

This would show a blue image.

Upvotes: 8

Related Questions