Monocular Camera Calibration using Projective Invariants

Abstract

Camera calibration is a crucial step to improve the accuracy of the images captured by optical devices. In this paper, we take advantage of projective geometry properties to select frames with quality control points in the data acquisition stage and, further on, perform an accurate camera calibration. The proposed method consists of four steps. Firstly, we select acceptable frames based on the position of the control points, later on we use projective invariants properties to find the optimal control points to perform an initial camera calibration using the camera calibration algorithm implemented in OpenCV. Finally, we perform an iterative process of control point refinement, projective invariants properties check and recalibration; until the results of the calibrations converge to a minimum defined threshold.