Improving the Condition Number to Achieve a Robust Camera Calibration Process

C. Ricolfe-Viala and A.-J. Sánchez-Salmerón (Spain)

Keywords

Camera calibration, Condition number, Homography estimation.

Abstract

Camera calibration is an important step in several processes of quantitative measurements or 3D modeling. Usually, the process consists of estimating camera parameters using an image of a calibration pattern. Correspondences between the image and the pattern points are arranged in a matrix to estimate camera parameters. The condition of this matrix represents an index of robustness of the calibration process. The aim of this paper is to define optimal conditions under which this matrix is best conditioned. It depends on the setup of the planar pattern, the location of the camera in the scene and its internal features. Once this conditions are defined several processes of camera calibration arise which guaranties the robustness of the estimation process.

Important Links:



Go Back