In this thesis, different approaches to calibrate an optical microscope will be presented. State-of-the-art algorithms are implemented and improvements are done to get more stable results and make the calibration process more flexible. Algorithms are developed to calibrate some of the camera parameters in advance in order to achieve better end results for the calibration process. In addition, techniques will be discussed on how to reduce errors produced by noise on a given input data set so that correct parameters for a specific camera model can be calibrated. These algorithms are evaluated and compared on real and syntetic data sets.