Python: grayscale image to RGB image with provided palette

Question

I am stuck with a problem.

I want to convert many greyscale images to RGBA with my dictionary palette where greyScale value = (r,g,b) (adding transparency 0 where the pixel is 0).

I wrote some code but it takes ~0.3sec to end one transformation. Is there some faster way to get better performance?

Thanks <3

Answer 1

The performance depends on your image resolution. But you can use Numpy like this for the case.

import numpy as np
from PIL import Image

def grayscale_to_rgba_optimized(image, palette):
    """Converts a grayscale image to RGBA using a dictionary palette efficiently.

    Args:
        image: A 2D grayscale image array.
        palette: A dictionary mapping grayscale values to RGBA tuples.

    Returns:
        A 3D RGBA image array.
    """

    lookup_array = np.array([palette.get(i, (0, 0, 0, 0)) for i in range(256)])
    rgba_image = lookup_array[image]
    return rgba_image

colors = {
    -35: (243, 243, 243),
    -30: (207, 207, 207),
    ......
    90: (77, 77, 77)
}

gray_img = Image.open('Your_Image_Path.png').convert('L')
gray_array = np.array(gray_img)

rgba_image = grayscale_to_rgba_optimized(gray_array, colors)

# Save the result as a PNG image (transparency will be preserved)
result_image = Image.fromarray(rgba_image, 'RGBA')
result_image.save("output_image.png")