机器视觉检测相比传统的坐标测量机具有成本低廉、检测速度快的优点。对以往的远心镜头的标定方法进行了对比研究, 提出了一种基于一组顺序图像并且简化外部参数的标定方法, 建立了远心镜头的模型, 考虑了镜头的三种畸变情况, 把复杂参数的矩阵求解转化为镜头参数的非线性最优化求解问题。通过实验标定测试, 远心镜头最佳测量景深为对焦点前后-9~8mm, 保证系统处于6 μm以内的精度范围, 获得的畸变系数等参数项可以为视觉检测系统提供重要的变换依据, 为建立高精度视觉测量系统建立了应用基础, 具有工业应用价值。

Compared with traditional coordinate measuring machine (CMM), visual inspection has special advantages such as lower cost and higher detection speed. The calibration method for telecentric lens is studied, and a simple calibration method based on ordered images sequence is proposed. The mathematic model of telecentric lens is established, and the lens distortion is solved by a non-linear equation. Lens distortion is taken into consideration to improve the calibration accuracy, the center point and internal and external parameters are also available. The telecentric lens calibration is solved through experiments with a higher calibration accuracy. The telemetry depth of telecentric lens used is about -9mm to 8mm before and after the focus plane, which can guarantee the accuracy range within 6μm. The obtained distortion coefficients can provide an important transformation basis for the visual inspection system. The high-precision visual measurement system will be established in next step with a vital industrial application value.