To capture extended depth of field video, we propose to move the detector at a constant speed, forward one frame, back the next and so on. This is possible, because the blur kernel is invariant to the direction of motion.
Back to project page
