User:Peter.fasogbon89/sandbox

Calibration problems can be cast as determining the transformation between two reference frames. Depending on the nature of the available measures and where the reference frames are attached we have four calibration or orientation problems1 : Relative orientation is the problem of determining the position and attitude of one perspective camera with respect to another camera from correspondences between points in 2-D images. Absolute orientation is the problem of aligning two sets of points, whose 3-D locations have been measured (or reconstructed) in two different reference frames. Exterior orientation is the problem of determining the position and attitude of a perspective camera from correspondences between 3-D points and their 2-D images.

Interior orientation is the problem of determining the (affine) transformation of the image plane from normalized camera coordinates to pixel coordinates. The solution to the interior and exterior orientation problems provides an overall solution to the camera calibration problem (or resection) that relates the position of pixels in the image to 3-D points.

THis explanations are from photogrametry

Camera calibration/resectioning is a major branch of computer vision field. They are used in major vision problems such as robot navigation, stereovision, structure from motion etc [Tsai86, Faugeras93, Fitzgibbon98, Kumar94, Wilson94, Heikkila97, Pollefeys00, Zhang00, Debevec01, Kurazume02]. The idea of camera resectioning is to estimate the parameters of an un-calibrated camera using a mathematical model. The paremeters to estimate are the external parameters i.e the position and orientation of camera related to a world coordinate system, and the internal parameters which comprises of the principal point or image centre, focal length and distortion coefficients (See the figure below). One of the most used camera calibration techniques is the method proposed by Tsai [Tsai86], and Zhang.

Unlike Zhang that involves the use planar object for calibration (planr object means one axis betweeen X, Y or Z) is equal to zero. Tsai usesa 3D calibration target.