The calibration of simple handheld instruments is often more expensive than the price of a new device. Therefore, the amount of manual labour is kept at a minimum in order to keep the price of calibration at a tolerable level. This also means that only a few points of e.g. a length scale can be checked. By using automatic machine vision based systems, the calibration of measurement instruments can be done faster and more thoroughly. In order to study the possibilities of machine vision automation for volume calibration tasks a set-up for micrometer calibration was constructed at Centre for Metrology and Accreditation (MIKES). With the developed automated machine vision system it is possible to check hundreds of points on the scale of a micrometer, giving new insight into error sources of the micrometer screw. The attained uncertainty is at the same level as calibration with gauge blocks according to ISO 3611.