基于圆环点的亚像素摄像机自标定方法  

Sub-pixel camera self-calibration approach based on circle-points

作　　者：杨根齐[1] 汤宝平[1] 蒋恒恒[1] 

机构地区：[1]重庆大学机械传动国家重点实验室,重庆400030

出　　处：《中国测试》2009年第3期107-109,共3页China Measurement & Test

基　　金：霍英东教育基金会11届青年教师基金(111057);教育部留学回国人员启动基金(教外司留(2008-5);第四十三批博士后科学基金(20080430749)

摘　　要：针对现有的摄像机自标定算法抗噪声性能差等问题,分析了基于圆环点的摄像机自标定方法的原理,对现有方法作出了改进。充分利用Harris算子和Fostner算子的优点设计了一种亚像素的角点检测算法,提取的角点位置可精确到亚像素级;同时引入了等效圆心的概念,最大限度地降低了噪声干扰。然后充分利用射影几何调和共轭的理论求取摄像机的内参数。实验结果表明,该标定方法原理简单,标定结果鲁棒性较好且标定精度较高,可应用于高精度的视觉测量。In view of the poor anti-noise performance of traditional camera self-calibration, the principle of camera self-calibration has been analyzed based on circle-point, and some improvements have been made based on the traditional methods. Making full use of the merits of harris operator and fostner operator, a novel subpixel comer suspecting algorithm was designed, and the location accuracy using this algorithm could come to sub-pixel. In order to decrease the effect of noise at higher degree, the equivalent center of the circle was imported. Then, it made full use of the theory of harmonic conjugate in projective geometry to calculate .the inner parameters of camera. The experimental results demonstrate this method does very well on the theorefie simplicity, the robustness and veracity of the result. And it can be applied to high-accurate vision measurement.

关 键 词：摄像机自标定 圆环点 等效圆心 HARRIS算子 Fostner算子 

分 类 号：TP334.22[自动化与计算机技术—计算机系统结构] TP751[自动化与计算机技术—计算机科学与技术]