A novel camera calibration method based on circle projection model

Abstract

Camera calibration is one of the important factors affecting measurement accuracy in the field of visual measurement, and the camera calibration method with high precision and strong robustness is the prerequisite for ensuring high-precision visual measurement. Among the commonly used calibration methods, target-based calibration methods have high accuracy and good flexibility. However, the feature extraction of checkerboard targets is prone to noise interference. Circle targets have higher feature extraction accuracy, but will introduce eccentricity errors in the center of the circle projection. In order to eliminate the eccentricity error of circle targets, this paper proposes a camera calibration method based on the circle projection model, which does not rely on the central projection coordinates, but directly constructs the mapping between the circle equation on the target and the elliptic equation in the camera image, and solves the optimal mapping parameters through nonlinear optimization algorithm. Through simulation and experiments, it has been proven that the calibration method based on the circle projection model has high calibration accuracy and anti-noise interference ability, providing a guarantee for high-precision visual measurement.