为提高标定精度,提出一种基于直线空间变换的光平面标定方法。首先利用互相关模板与5点滑动平均法提取激光条纹中心,然后采用正交回归法拟合图像中的光条直线方程。通过平面单应性变换获得靶标面光条直线方程,进一步再将靶标坐标系中的直线方程转换成Plücker矩阵形式。根据位姿转换关系得到相机坐标系下的直线方程,并建立超静定线面共面约束方程组,使用奇异值分解(SVD)求解光平面方程参数。所提方法测量的标准台阶块长度方均根(RMS)为0.065mm,平均误差小于0.030mm,圆柱直径的测量平均误差与RMS小于0.050mm。结果表明,利用光条自身整体信息拟合光平面,所提方法可实现较高的标定精度且标定过程简单,同时避免对每个特征点进行单独求解。

To improve calibration accuracy, this paper proposes a light plane calibration method based on a line transformation in space. First, a cross-correlation template and a 5-point moving-average method are used to extract the center of the laser stripe, and the linear regression equation of the light bar in the images is then fitted by the orthogonal regression method. A straight line equation for the target surface light bar is obtained by a plane homography transformation, and the straight line equation in the target coordinate system is converted to Plücker matrix form. According to the pose transformation relationship, the straight line equations in the camera coordinate system are obtained and statically indeterminate line-plane co-planar constraint equations are established. A singular value decomposition(SVD) is used to solve the parameters of the light plane equation. The root mean square (RMS) of the standard step block length obtained by the proposed method is 0.065mm, and the average error is less than 0.030mm. The measured average error of cylinder diameter and RMS are less than 0.050mm. The results show that the proposed method uses the overall information of the light bar itself to fit the light plane, which can achieve higher calibration accuracy. This represents a simple calibration process that avoids the separate solution of each feature point.