摘要
在研究传统的摄像机标定方法的基础上,提出了一种基于十字同心圆标定模板的摄像机标定新方法。该方法通过从2个或2个以上不同的方位上拍摄带有十字同心圆模板,采集具有标定特征的图像,从而完成摄像机的标定。这种基于曲线拟合的标定方法,与传统的提取参考点的标定方法不同,采集的参考点坐标相对准确,计算的摄像机标定结果具有很高的准确性;同时该方法原理简单,实现方便,具有很好的可操作性。通过实验验证表明,该摄像机标定方法具有较高的准确度、较小的误差率,可以获得较为满意的标定参数。
By studying traditional methods for camera calibration, a new method based on concentric circles with cross lines is proposed. Camera calibration is completed by taking pictures from two or more different locations and acquiring the images with calibration characteristic. Different from the traditional method of extracting reference points, the new method based on curve fitting can get precise coordinates of reference points, which assures the accuracy of the results of camera calibartion. Meanwhile this method owns the characteristics of simplicity in theory, ease in implementation and good maneuverability. Experimental results show that the method can obtain high accuracy, low margin of error and better effect of calibration.
出处
《半导体光电》
CAS
CSCD
北大核心
2012年第4期582-587,共6页
Semiconductor Optoelectronics
基金
哈尔滨工业大学科研创新基金项目(HIT.NSRIF.2009023
HIT.NSRIF.2011070)
机器人技术与系统国家重点实验室项目(哈尔滨工业大学)
关键词
摄像机标定
同心圆
曲线拟合
可操作性
camera calibration
concentric circles
curve fitting
maneuverability
