任意位姿平面靶标实现立体视觉传感器标定

Calibrating Stereo Visual Sensor with Free-position Planar Pattern

提出一种基于任意位姿平面靶标的立体视觉传感器标定新方法。利用平面标定参照物,并允许其在测量空间内自由移动,结合镜头畸变的摄像机标定数学模型,实现传感器中摄像机标定;在立体视觉传感器三维测量模型基础上,引入特征点约束优化传感器结构参数,同时实现了传感器现场标定。该方法降低了标定设备的成本,灵活方便,切实可行。实验结果表明,该方法标定精度高,已标定传感器测量空间距离的相对误差优于0.3%。

A novel calibration method to determine all the primitive parameters of the stereo visual sensor with free-position planar pattern is proposed. In this method, without restriction to the moving of the planar reference object ,all the camera parameters can be readily calibrated with the mathematical model of lens distortion. And based on the 3-D measurement model of stereo-visual sensor, the parameters of sensor can be optimized with the feature points constraint simultaneously. The proposed approach greatly reduces the cost of the calibration equipment,and it is flexible and practical in the visual measurement. It is shown that this method has high precision by experiment,and the measured relative error of space length excels 0.3 %.