There are PixelDepth / SceneDepth to get distance to the objects at point of currently rendered pixel. What about to get an static mesh object by current pixel? I would like to render separate frame with a map of objects visible from camera perspective (to classify the objects).