In most computer vision tasks camera calibration is the first step before applying further algorithms like stereo matching or optical flow estimation. The main aspect of camera calibration is the determination of the geometric mapping of world points into the camera image. For a single camera this mapping can be described with a camera model and a set of intrinsic parameters. For stereo or multi-camera setups the calibration includes a description of the position and orientation of each camera towards a given reference (extrinsic parameters).
In our research we develop calibration routines for multi-camera setups with non-overlapping fields of view as well as special camera models for wide-angle optics.