Monocular pose acquisition
Active and passive machine vision applications in automated production have been explored. Among them are visual inspection, visual feedback control and navigation guidance. All require that mounting position of the vision camera and effective focal length be specified. The need for an efficient method which identifies on-site, effective focal length and mounting position of the vision camera is urgent. The thesis explores the monocular pose acquisition method. The vision camera sees well-structured object, which can be either a cube or a 2-D square plotted on a piece of paper of which the edge length is known. The monocular pose acquisition method determines both effective focal length and mounting position of the vision camera using the monocular image of the reference object gathered by the vision camera.