Cartesian Coordinates of model for each pixel in photo

[jamiemccoll1@gmail.com]

The final product I am aiming for is an array where each pixel consists of an x,y,z float of where that point in the image is found on the model. Similar to what renderDepth does for the depth of a given image, I hope to get 3D position instead. This to be used in conjunction with the original image such that a click on a photo can give the x,y,z position of that point in 3D space.

Initial thought is to simply use Camera.unproject() for each pixel, but from reading other posts it sounds like this is not correct.

Any help or hints will be much appreciated!