一种基于共线特征点的线阵相机内参标定方法  

A novel method for calibrating intrinsic parameters of linear array cameras based on collinear feature points

作　　者：王澜[1] 孙博[1] 隆昌宇[1] 邹剑[1,2] 邾继贵[1] 

机构地区：[1]天津大学精密测试技术及仪器国家重点实验室,天津300072 [2]军械工程学院光学教研室,河北石家庄050003

出　　处：《红外与激光工程》2015年第6期1878-1883,共6页Infrared and Laser Engineering

基　　金：国家杰出青年科学基金(51225505);"十二五"国家"863"计划(2012AA041205)

摘　　要：提出一种利用共线的标定特征点确定线阵相机内参的方法。所有标定特征点均在线阵相机的视平面内。首先,标定过程中对标定特征点逐一成像,直接得到标定特征点和其对应像点,解决了空间标定点与像点的对应问题;其次,通过数学建模得到线阵相机的成像模型,联立多个位置处的成像模型解算出相机内参;最后,对影响线阵相机标定的因素进行分析及实验验证。理论分析和实验结果表明,文中的线阵相机标定方法简单灵活,无需制作精密的靶标,可获得大量标定点,提高了线阵相机标定的准确性和稳定性,实验得到成像特征点重投影像点位置偏差均方根为0.37 pixel。A novel method was proposed that can be used for determining intrinsic parameters of linear array cameras based on collinear calibration feature points. All calibration feature points were on the view plane of the linear array camera. Firstly, during the camera calibration the feature point was captured by the camera one by one, the corresponding image point and the spatial feature point were obtained directly, which solved the matching problem of the spatial feature point and image point. Secondly, a mathematical model was established to obtain the imaging model of the linear array camera, then use the simultaneous imaging models of multiple locations was used to calculate the camera intrinsic parameters; Finally, the impacts on the calibration result were analyzed and experimented. Result reveals that the proposed linear array camera calibration method is not only convenient and flexible, but also sophisticated calibration pattern is not required, and a lot of calibration points are obtained which improving the accuracy and stability of the line scan camera calibration. The root mean square error of re-projected points is less than 0.37 pixel.

关 键 词：线阵相机 共线特征点 成像模型 标定误差 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]