基于OpenCV和本质矩阵的双目立体视觉摄像机标定方法

摘要

为克服传统标定过程复杂、不适用于现场操作等弱点，采用基于 OpenCV 和本质矩阵的双目视觉摄像机标定方法。该方法利用OpenCV编程在线标定摄像机的内部参数，通过图像对应特征点匹配，结合极线约束方程，线性求解出本质矩阵，从而快速地得到摄像机的外部参数。实验结果表明，该标定算法简单快速，可适用于户外场合的标定。%　　Most traditional approaches for stereo sensors calibration often require accurate calibration objects and complex procedures, which limit their capability in out-door work. The paper of binocular vision camera calibration method works on OpenCV and essential matrix. The method uses the program of OpenCV to calibrate the parameters of camera's internal on line. Equation of the essential matrix is formed by finding matched points in the image-pair, combined with the epipolar constraint and solved linearly. The internal parameters of the camera are rapidly got. The experimental results show that the calibration algorithm is simple and fast, and can apply to calibrate the occasion of out-door.