Product pattern-based camera calibration for microrobotics

Traditional macro camera calibration techniques use multiple images at multiple orientations to obtain the camera's internal and external calibration parameters. Microscopic camera calibration differs in this by limiting to only one image at fixed orientation (parallel to the image plane) and requiring a special calibration pattern. We propose a method to use the repetitive product pattern as calibration object. A parametric model of an optical microscope validates this approach by comparing the results from the product pattern with an industrial high-accuracy calibration pattern.