基于虚拟标定场的数码相机内参数标定方法  

New Camera Calibration Method Based on Virtual Calibration Field

作　　者：吴少平[1] 张爱武[1] 臧克[1] 

机构地区：[1]首都师范大学三维信息获取与应用教育部重点实验室,北京100037

出　　处：《系统仿真学报》2006年第z2期424-426,共3页Journal of System Simulation

基　　金：国家科技部十五科技攻关项目(2004BA810B);北京市属市管高等学校"学术创新团队计划"项目;北京市教委科技发展计划项目(KM200610028012)。

摘　　要：建筑物表面纹理数据获取与真实纹理映射一直是数字城市亟待解决的难点问题,也是真实感图形学研究的重点内容。结合相机成像理论,提出一种基于虚拟标定场的数码相机标定方法,通过一组图像构建虚拟标定场,解算数码相机内部参数,快速自动纠正图像畸变,为实现真实纹理映射奠定基础。并对该方法获得的标定参数进行了实验,证明满足纹理映射的要求。In order to acquire high quality texture mapping data, which is an important challenge for constructing Cyber-City, using camera imaging theories, a new camera calibration method was proposed. A virtual calibration field was built through a serial of images, and then the interior parameters and distortion coefficients of camera were fast calculated from the serial of images. For all the work is done in or by the scanner coordinate system, the total-station is not needed. Experiments show the method is robust.

关 键 词：三维模型 三维可视化 相机标定 畸变 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]