0
$\begingroup$

I want to do a test for image brightness in the rendered image (bpy.data.images["Render Result"]). I would like to be able to see if the image is dark or light. An example output of the code could be The rendered image is: light

$\endgroup$
2
1
$\begingroup$

You can access the pixel data directly through the Image.pixels property. It is a list of repeated RGBA values. To access a single pixel, use this:

img = bpy.data.images["Render Result"]
value = img.pixels[(y * img.width + x) * 4 + channel]

where channel is 0 for red, 1 for green, 2 for blue and 3 for transparency.

Now, image brightness is not a clearly defined thing. You can, for example, calculate the average pixel value over the whole image:

r, g, b = [sum(img.pixels[:][i::4]) / len(img.pixels) for i in range(3)]

That was "average" as in "arithmetic mean". Another possible interpretation of image brightness is a quantile, say, the 95th:

import numpy as np
r, g, b = [np.quantile(img.pixels[:][i::4], 0.95) for i in range(3)]

The specific choice probably depends on the kind of images you are processing.

$\endgroup$
3
  • 1
    $\begingroup$ This just gives me 1.0 every time. $\endgroup$ – John Roper Nov 27 '16 at 19:19
  • 2
    $\begingroup$ The maximum value is often 1, and it means that at least one pixel of the whole is bright. If you want a more specific measure of brightness, be more specific in your question. $\endgroup$ – emu Nov 28 '16 at 14:03
  • 1
    $\begingroup$ So how to I calculate image brightness from this? $\endgroup$ – John Roper Nov 30 '16 at 1:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.