自动化学报2019，Vol.45Issue(10)：1995-2000,6.DOI:10.16383/j.aas.c180596
基于两个共面圆的无匹配相机位姿计算
Efficient Camera Localization From Two Coplanar Circles Without Matching
摘要
Abstract
Camera localization attracts strong interests among both academic and industrial communities, of which the methods based on the planar markers are greatly easy to implement. However, once one matching between markers and their images is wrong, the computed camera pose will be unstable. In this paper, a new method for camera pose computation is proposed from the designed two coplanar circles. Matching among multiple points is not needed. The solutions are analytical and lightweight. Moreover, due to that the used image features are edges, the method is still successful to rapid camera motion and image blur. These increase the robustness of camera pose computation and decrease the computational cost greatly. Experiments including comparisons validate the effectiveness of the proposed method.关键词
相机标定/相机定位/相机位姿跟踪/圆形图像Key words
Camera calibration/camera pose computation/camera pose tracking/image of circles引用本文复制引用
范蓉蓉1,2,唐付林2,3,刘青山1..基于两个共面圆的无匹配相机位姿计算[J].自动化学报,2019,45(10):1995-2000,6.基金项目
国家自然科学基金(61836015, 61572499) 资助 （61836015, 61572499）
