双摄像机靶标成像视觉坐标测量方法研究

Research on Probe Imaging Vision Coordinate Measurement with Binocular Cameras

提出了一种利用手持式靶标实现双目视觉三维坐标测量的新方法.以靶标作为成像目标,利用靶标与被测面接触,通过分析靶标上已知特征点的成像变化,实现与靶标接触点的三维坐标测量.这一测量新方法,可以实现视觉系统外部参数测量中的自标定,并能实现遮挡点的测量.研究中利用空间透视变换建立了这一方法的非线性测量模型,介绍了基于Newton非线性迭代的测量方程的求解方法.以矩阵分析理论为基础,讨论了影响测量方程求解的因素,并提出了方法中重要部件手握式靶标上特征点分布的基本约束条件.通过数值计算分析了图像处理误差、有效焦距和基线标定误差对测量方法的影响.实验结果验证了理论分析的正确性和应用的可行性.

A novel three dimensional measurement technique using dual cameras combined with a hand-held probe is presented. By the imaging changes of known points on the probe, the 3D coordinates of the contacted points between the probe and measured surface are tested. The new approach can carry out the self-calibration of extrinsic parameters. It can be also used to measure these occluded points. Nonlinear measurement model is constructed by perspective transform. The Newton iteration is used to solve the measurement model. Based on the matrix analysis, the influential factors on the model solving are discussed and the restriction conditions to the distribution of the known points on the probe are proposed. By numerical calculation, the effect on the errors due to image processing, calibration of effective focal length and basic line is discussed. The experimental results prove the validity and feasibility of the approaches in theory and application.