针对深度空间小视角摄像机标定, 建立基于立体靶标的仿真分析方法, 深入分析了靶点的空间分布各因素对摄像机内参数标定精度的影响, 提出在摄像机标定之前合理设计靶点的数目及位姿分布的思想。分析表明, 靶点的空间分布范围对标定精度的影响最大, 靶点的数目对精度影响较小; 从最近测量距离开始布置靶点, 可提高标定精度。提出一种结合感兴趣区域和灰度矩亚像素边缘检测的图像光斑中心定位的算法, 有效保证了定位精度和运行效率。构建虚拟立体靶标对摄像机内参数进行了快速标定, 参数标定精度高, 具有实际应用价值。该方法也适用于无需靶标拼接的摄像机标定。

Aiming at the calibration of camera with a narrow field of view for deep space, the simulation analysis method based on stereo target is built. The spatial distribution influence of the target points on the accuracy of camera calibration is analyzed thoroughly. The number and position distribution of the target points are reasonably designed before camera calibration. The results show that spatial distribution scope of the target points has the greatest influence on the calibration accuracy while the number of target points has little influence. The calibration accuracy can be enhanced by arranging target points from the nearest measuring distance. A center location algorithm combining the region of interest with sub-pixel edge detection by the gray moment is proposed to locate the image target spot. The location accuracy and the processing efficiency of the image algorithm are improved effectively. The virtual stereo target is arranged to calibrate the intrinsic parameter of a camera quickly. The calibration precision is high and the method has practical application value. This method is also applicable to camera calibration without target stitching.