中文摘要：

在摄像机标定过程中，球形靶标对图像数量、拍摄角度要求小，能够适应遮挡环境，在使用中具有明显优势-在分析球投影模型几何性质的基础上，提出一种利用球靶标分步标定摄像机内部参数的方法-该方法通过拍摄一幅包含空间中2个不同位置球体的图像，得到图像中的2条投影二次曲线；利用2条投影二次曲线的对称轴和公切线计算球心投影点，通过投影二次曲线的对称轴确定图像主点坐标，根据球心投影点和图像主点的位置关系求解归一化焦距；利用整体优化算法得到内参标定结果-仿真数据实验分析了引入测量误差的主要因素，实物标定结果与平面方格靶标方法相比误差在5％之内，重复测量结果稳定-

英文摘要：

Camera calibration with spheres has little demand on image number, camera shooting angle and high adaptivity to occlusion, which has an obvious advantage in applications- A novel camera calibration method was proposed based on geometric feature of sphere projection model- This algorithm used two projec- tion conics of spheres and solved the intrinsic parameters of the camera- The projection conics were obtained from image of two spheres at different positions- The symmetry axis and common tangent of the projection con- ics were used to solve the principal point and the projections of sphere centers- Then normalized focal length was computed based on geometric relationship of the principal point and projection of sphere center- Finally intrinsic parameters were obtained using optimization algorithm- Simulation data experiments analyze the main error factors and real data results show that the proposed method is reliable and the error is less than 5% com- paring with planar target method-