I want to say that it isn't feasible, and as an example reference the demo picture with the bicycle wheel in the extreme foreground, where you can kind of see "through" the spokes when focusing on things in the distance …

but

If each pixel in the image is clickable and focuses around that pixel, what would happen if you clicked on each pixel and then copied that one pixel? I'm almost inclined to write an AutoIt script to do exactly that.