基于双目视觉的回转类零件测量方法研究
Positioning Method of Rotation Parts Based on Stereo Vision
基于机器视觉的测量技术目前广泛应用于自动化生产过程中的各个流程，与技术较为成熟的基于单目相机的非接触测量相比，双目视觉可以获取更多目标信息，有较高的研究价值。本文为探索双目视觉技术在非接触测量中应用，提出了一种用于获取回转类零件的位置和尺寸的测量方法，利用其成像特征和数字图像处理技术，实现了较为快速的定位与测量。在对目标物图像完成边缘检测后，根据轮廓中点的数目对边缘检测所得向量中的边缘特征进行筛选后拟合椭圆，可以有效地排除噪音干扰，实现定位。在测量结果与传统的三维重建算法所得结果偏差较小的基础上，大幅减少了计算量与用时。
In the process of non-contact measurement of rotation parts, the circular surface feature can be used for faster positioning and measurement. This paper proposes a stereo vision-based rotation part positioning method. According to the number of points in the contour, the edge features in the vector detected by edge detection are filtered to fit the ellipse, which can effectively eliminate noise interference, and achieve positioning. Measure results On the basis of less deviation from the results obtained by the traditional 3D reconstruction algorithm, the calculation amount and time are greatly reduced.
徐晓慧、程亚飞
机械零件、传动装置计量学自动化技术、自动化技术设备
双目视觉相机标定边缘检测椭圆拟合最小二乘法
Binocular visionCamera calibrationEdge detectionEllipse fittingLeast squares method
徐晓慧,程亚飞.基于双目视觉的回转类零件测量方法研究[EB/OL].(2020-04-15)[2025-08-16].http://www.paper.edu.cn/releasepaper/content/202004-116.点此复制
评论