Research On Camera Calibration Based On Two Coplanar Circles

As the development of computer technology, computer vision has been greatly applied in the field of mapping,industrial control,navigation,military,robot hand-eye system and so on. As an important step of spatial vision, camera calibration is the precondition and foundation of obtain 3D information by images. So in the recent years the research about camera calibration is very dynamic.In this paper, we summarize the achievement of domestic and abroad study, and focusing on the methods of camera calibration. Calculating the parameters of camera by structuring camera models, and then building the relationship of 3D objects and 2D images, laying the foundation of 3D reconstruction. The major elements of this paper is about calibration target designing and the detection of feature points. The specific research contents can be summarized as follows:1. Proposed an calibration method based on the target of two coplanar circles. The method selects two coplanar circle as calibration targets, getting the position of the feature points by the geometrical relationship of the two coplanar circles. It does not need any manual operation and can ascertain the position of the feature points automatically, then we can calculate all the camera parameter and finish camera calibration.2. Studied about feature points extraction. Aiming at the method proposed in this paper, by simulation experiment we compared the results of edge detection, and then chosen the Canny operator. After edge detection, we calculated the ellipse’s parameters by least square method,and then get the position of feature points.3. Finished camera calibration in the laboratory and obtained parameters of the camera. Designed the analysis and comparison experiment,used the method proposed in this paper and Zhang’s method calibrated the same camera. By comparing the result proved that the method proposed in this paper is feasible and effective.