How to get the worldspace location of a rendered pixel from Python

Question

I need a way to find the underlying 3D coordinates for each pixel of the camera in Python.

I to create a function like:

GetCoordinatesForPixel(xPixel, yPixel)

which returns a (X, Y, Z) vector.

I found the following threads:

The latter solution works in that it procudes a render. However, I do not see a way to access those values in Python.

Did you look at this post blender.stackexchange.com/questions/7106/…? You will have to play around with it a little, but this might be headed in the right direction. — BlendingJake, Nov 5 '14 at 14:29
Once for the actual image I want to estimate on and once to get the ground truth values. @BlendingJake I tried to work with bpy_extras.view3d_utils but so far I am not really sure how to handle the arguments. But I may be able to figure it out:) — user1033657, Nov 6 '14 at 14:09
@user1033657 I wasn't able to find much info on it, but just keep playing and you'll probably get it. — BlendingJake, Nov 6 '14 at 14:12
re-opening, @user1033657 would like to edit the question (to make it not a duplicate) — ideasman42, Dec 15 '14 at 17:53

ideasman42 · Accepted Answer · 2014-12-15 18:08:27Z

Blender's Python API currently doesn't expose access to the pixel value of render passes.

(You can access image pixels but renders's are a more complex structure).

However since you mention you need to store the values in a file, just save the OpenEXR file, this will store the values in a file.

1 Answer 1