提出一种新的基于并联机构的摄像机线性标定方法.首先推导出考虑一阶径向畸变针孔摄像机模型的完整表达式;然后巧妙利用六自由度并联机器人机构动平台作为二维平面模板的载体,构造虚拟三维靶标;最后,在确定一阶径向畸变系数和主点坐标的基础上,利用推导出的一阶径向畸变针孔模型表达式并结合虚拟三维靶标信息一次性线性求解摄像机其他内外参数.该方法可保证摄像机在固定之后进行现场标定,使摄像机的标定状态和工作状态完全一致,便于视觉并联机器人相关问题的研究.仿真和真实实验结果表明,该方法简单快捷,不仅避免了三维靶标的制作问题,而且避免了非线性优化的不稳定性,能满足实际需要.%A novel linear approach for camera calibration based on parallel mechanism was proposed. First, the integrated expression of camera model with the first-order radial distortion was educed. Then a virtual 3D target was constructed by using the motion platform of the 6-DOF parallel mechanism as the carrier of 2D planar template. After the first-order radial distortion parameter and the principal point coordinates were solved, the other parameters were computed linearly using the foregoing camera model and the information of virtual 3D target. This approach can make the calibration state consistent with the work state,it eliminates 3D target's facture and the instability of nonlinear optimization. Experiments with simulated data as well as real image show that the approach is feasible and efficient.
展开▼
