Many approaches exist for this type of calibration. Most of these methods consist of a two step procedure where a calibration is obtained first for a simplified (linear) model and then a more complex model, taking distortions into account, is fitted to the measurements. The difference between the methods mainly lies in the type of calibration object that is expected (e.g. planar or not) or the complexity of the camera model that is used. Some existing techniques are Faugeras and Toscani [40], Weng, Cohen and Herniou [215], Tsai [202,203] (see also the implementation by Willson [219]) and Lenz and Tsai [98].