Shape and Pose Recovery of Novel Objects Using Three Images from a Monocular Camera in an Eye-In-Hand Configuration