论文部分内容阅读
提出基于全局位姿评估的条纹反向视觉测量。利用带测头的摄像机拍摄二维正弦条纹,根据傅里叶分析提取的平面特征点全局评估摄像机位姿,进而测量测头的球心坐标。摄像机的位姿评估是条纹反向视觉测量的关键技术,在测头标定和坐标测量中均需要进行位姿评估。位姿评估的误差函数包括重投影误差和物空间误差两类,两者在基于平面的位姿评估中均存在两个局部极小值,采用全局位姿评估算法可以避免误差函数陷入局部极小以高精度获取摄像机位姿,完成测头标定和坐标测量。实验结果表明,该方法可以高精度测量物体的三维坐标。
The stripe inversion visual measurement based on global pose evaluation is proposed. A two-dimensional sine fringe is taken by using a camera with a touch probe, and a global evaluation of the camera pose is performed according to the planar feature points extracted by Fourier analysis to measure the ball center coordinates of the touch probe. Camera posture evaluation is the key technology of the stripe inversion visual measurement, which needs attitude evaluation in the calibration of the probe and the coordinate measurement. The error function of posture evaluation includes two kinds of re-projection error and object space error. Both of them have two local minima in plane pose-based evaluation. The global pose estimation algorithm can avoid the error function from falling into local minima Get the camera pose with high precision, complete the probe calibration and coordinate measurement. Experimental results show that this method can measure the three-dimensional coordinates of objects with high accuracy.