基于标准杆的双目视觉传感器标定技术研究

目前制造业对大范围精密测量技术要求越来越高,为了克服传统双目立体视觉标定过程中测量场难实现、标定精度低等缺点,提出了一种较灵活的基于一维标准杆的双目视觉传感器的标定方法。该方法首先利用正方形模板和带有四个特征点的标准杆实现主点和畸变的标定,然后在测量场多次摆放标准杆,由两摄像机拍摄标准杆图像,利用标准杆上对应特征点之间的约束以及基本矩阵估计双目视觉传感器其余参数,最后光束法平差进行整体优化,消除累积误差。仿真实验和真实实验结果均表明该方法适用于大视场双目视觉传感器的标定,操作简单、标定效率高。

Research on Calibration of Stereo Vision Sensor Based on Standard Bar

The more accurate of large scope precise measurement is increasingly required in manufacturing industry,this paper presents a more flexible method based on one-dimensional standard bar for calibrating stereo vision sensor,which overcomes difficulty and low precision in factory in traditional binocular stereo vision sensor. Firstly, a square template and a standard bar with four points are used to calibrate principal point and aberration, then the standard bar is ran-domly placed in measurement space for several times and the images of the points on the standard bar are acquired by the cameras,the other parameters are estimated based on the foundational matrix and the constraint of the known dis-tance between the points on the target, lastly, the bundle adjustment method is applied to optimize all the parameters of the cameras and calibration points to reduce the accumulated error. The results of simulation and physical experiment show that the proposed calibration method is efficient and easy to operate, and this method is suitable for the calibra-tion of stereo vision sensor with large field of view.