中文摘要：

立体相机的标定是一个精确求解各个相机内参数以及相机之间关系参数的过程。它是三维重建的基础，其标定精度的好坏直接影响立体重建的结果。为此提出了一种使用校正误差作为代价函数的立体相机标定算法。该算法首先使用传统的基于重投影误差的方法对单个相机的内参数进行标定，然后利用校正误差完成对相机之间关系参数的标定求解。由于校正误差的计算只与相机内参数以及关系参数有关，可以避免在标定过程中使用难以精确标定的相机外参数。实验结果表明本算法能够有效的提高立体相机标定的精度。

英文摘要：

The process of stereo calibration is to determine the intrinsic parameters of each camera and the relative parameters between stereo cameras. It is an important stage in stereo reconstruction, the quality of calibration has a direct impact on the quality of recon- structed results. A new stereo calibration algorithm based on rectification error is presented here. The algorithm f＇trst calibrates intrin- sic parameters of each camera by traditional method based on reprojection error, then, solves the relative parameters by using rectifi- cation error. Since rectification error can be calculated only by intrinsic parameters and relative parameters of the stereo cameras, the accuracy of stereo calibration can be improved by avoiding using camera extrinsic parameters which are difficult to be calibrated accu- rately. Experiments show the efficiency of the proposed method to improve the accuracy of stereo calibration.