Camera Calibration Using Composed Cubic Splines

Abstract

Camera calibration is a process whereby the geometric characteristics of a specific camera are determined. It is performed so that the photograph obtained can be used to produce accurate measurements. This paper presents a simple and accurate model for extracting the optical characteristics (intrinsic parameters) as well as the 3D position and orientation of the camera (extrinsic parameters) based on image coordinate measurements for targets of known 3-D position. The proposed algorithm is divided into two major steps. First, the behaviour of a grid of projected lines is studied and the required local corrections are applied. Such corrections are calculated using Composed Cubic Splines. This function covers radial, decentering and prism distortions. It provides a more realistic model of distortion which is represented as a composed surface. Then, calibration parameters are estimated using a pinhole model based on the minimization of a linear criterion relating the 3D coordinates of the targets with their 2D image projections.