摘要
提出了一种新的基于二维靶标的球面模型鱼眼镜头标定方法。根据球面成像模型中圆弧上的点与图像坐标系像点之间的约束关系,利用平面靶上的一根直线上的点,初步估计出摄像头的内部参数。利用平面靶上棋盘格角点在世界坐标系上的坐标点与其在球面坐标系下坐标的映射关系,求出相应的外部参数的初始值。以这些角点在图像中的实际坐标与重投影后坐标之间的均方差为优化参数,进行非线性优化,求出内部参数,畸变系数和外部参数的精确解。通过实验验证,该方法能够快速的估计鱼眼镜头的内、外部参数的初始值,具有较高的精度,能够满足实际应用的需求。
A new calibration method based on the spherical model for fisheye lens using two-dimensionally calibration target is presented. According to the relationship between the point on arc and the image points for the spherical model, the internal parameters by the points in a straight line on the plane target are preliminary estimate. And the corresponding external parameters by the mapping between the world points and the points in arc on the sphere are calculated. The rooy- mean- square error of the image points and the reprojection points are taken as the optimization parameter and the exact solution of the internal parameters, the distortion coefficients and the external parameters are solved by the nonlinear optimization algorithm. The experiment presents that the proposed method can quickly estimate the initial values of the internal and external parameters for the fisheye lens with high-precision and meet the practical applications.
出处
《中国激光》
EI
CAS
CSCD
北大核心
2015年第5期226-233,共8页
Chinese Journal of Lasers
基金
国家自然科学基金仪器专项(61027014)
国家自然科学基金青年项目(11109052)
深圳市科技计划项目(JCYJ20140418091413574)
深圳市战略性新兴产业发展专项(JCYJ20120614085204873)
关键词
测量
摄像机标定
鱼眼镜头
非线性优化
measurement
camera calibration
fisheye lens
nonlinear optimization algorithm
