【作者】 刘广琦； 李德强； 郝颖明；

【Author】 Liu Guangqi~(1,2,3,4) Li Deqiang~(2,3,4) Hao Yingming~(2,3,4) (1 Graduate University,CAS,Beijing 100049;2 Shenyang Institute of Automation,CAS,Photoelectric Information Laboratory, Shenyang 110016;3 Key Laboratory of optoelectronic information Processing,CAS;4 Liaoning Province Key Laboratory of Image Understanding and Visual Computation)

【机构】 中国科学院研究生院； 中国科学院沈阳自动化研究所光电信息研究室； 中国科学院光电信息处理重点实验室； 辽宁省图像理解与视觉计算重点实验室；

【摘要】 本文基于Tsai的经典两步法提出了一种改进的摄像机两步标定法。该方法针对一阶径向畸变模型重新定义了摄像机的内外参数,将Tsai的经典两步法中参数之间的非线性关系转化为线性关系,并利用简化的直线搜索算法进行一维迭代求解。方法避免了多变量的非线性优化且不再需要设定寻优变量的初值,因此算法稳定,计算量小。更多还原

【Abstract】 Basing on Tsai’s classic two-stage approach for camera calibration,this paper proposes an improved two-stage approach.This approach based on the first order radial distortion model,redefines the internal and external parameters of camera to implement the linearization of the nonlinear equations about the parameters of camera in the Tsai’s classic two-stage approach and configure them out using a simplified straight line search algorithm.This approach has no need to solve the multivariable nonlinear optimization problem and to set the initial value of the optimization variables.So,the algorithm of this approach is stable and has low computational complexity.更多还原