双目立体视觉系统的非线性摄像机标定技术*

针对双目立体视觉的工业检测精度和实时性要求，采用LENZ畸变模型建立基于面阵CCD的双目成像几何模型，分析了成像模型的内外参数及针孔模型的局限性，提出改进的双目立体视觉系统摄像机两步标定方法。利用HALCON标定板及函数库平台，采用亚像素精度的边缘提取和椭圆拟合算法精确获取标定点，以建立像点与空间点的对应关系，并对标定参数进行非线性优化。实验证明该标定算法灵活准确，并具有良好的可扩展性。

Nonlinear camera calibration method of binocular stereo vision system

Under the requirements of industrial inspection accuracy and real time application, this paper brought forward an improved two-step camera calibration method for binocular stereo vision systems based on the binocular imaging geometry model of the area-scan CCD camera with the using of LENZ distortion model, analyzed of the parameters and the limitations of the pinhole model. Using HALCON planar calibration board which were circular targets of two-way array and functions library, obtained the calibration marks and determined the marks and their projections with the sub-pixel precise edge contour detection algorithm and fitting ellipse method. The camera parameters were nonlinear optimized during this process. The accuracy of these parameters and stability of the algorithm were validated through experiments and accuracy analysis. The calibration me-thod is flexible and good portability.