[1]许贤泽,钟明,刘盼盼,等.光纤棒机床自动定位与对准装置[J].郑州大学学报(工学版),2020,41(06):1-6.[doi:10.13705/j.issn.1671-6833.2020.06.012]
　Xu Xianze,Wang Xingyu,Liu Panpan,et al.Device for optical fiber preform machine tools automatic location and alignment[J].Journal of Zhengzhou University (Engineering Science),2020,41(06):1-6.[doi:10.13705/j.issn.1671-6833.2020.06.012]

点击复制

光纤棒机床自动定位与对准装置()

分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:

41卷

期数:

2020年06期

页码:

1-6

栏目:

出版日期:

2020-12-31

文章信息/Info

Title:

Device for optical fiber preform machine tools automatic location and alignment

作者:

许贤泽; 钟明; 刘盼盼; 王星宇

武汉大学电子信息学院,湖北武汉430072, 武汉大学电子信息学院,湖北武汉430072, 武汉大学电子信息学院,湖北武汉430072, 武汉大学电子信息学院,湖北武汉430072

Author(s):

Xu Xianze;  Wang Xingyu;  Liu Panpan;  wait.

Wuhan University School of Electronic Information, Hubei Wuhan 430072, School of Electronic Information, Wuhan University, Wuhan 430072, Hubei University, Wuhan University School of Electronic Information, Hubei Wuhan 430072, Wuhan University School of Electronic Information, Hubei Wuhan 430072 430072

关键词:

自动对准;  动态阈值;  边缘检测;  圆检测;  同轴检测

Keywords:

Automatic alignment;  dynamic threshold;  edge detection;  round detection;  coaxial detection

DOI:

10.13705/j.issn.1671-6833.2020.06.012

文献标志码:

A

摘要:

针对现有光纤棒沉积机床同轴调试中存在对准耗时长和精度低的不足，设计并实现了一套高精度、高稳定性的光纤棒机床自动定位与对准装置。首先，通过张正友标定法消除镜头畸变和其他误差；接着，采用高斯滤波、动态阈值法和canny边缘检测对提取准确的图像边缘，消除噪声和背景对后续处理的影响；最后使用新型弧分类圆形检测算法实现图像中被测面中心的精确定位，确定图像比例尺并实时显示偏差大小，指导调节机床卡盘位置，实现机床两端中心对准。将本文算法与几种常用圆检测方法进行比较，其平均误差为1.0664pixel，平均定位时间为244ms，优于其他检测算法。实验结果表明，该装置对机床对准的准确度和效率均有显著提升，在相机分辨率为500w像素时，系统检测误差小于0.1mm。满足机床同轴检测要求，为机床同轴调试提供了快速有效的方法。

Abstract:

Aiming at the shortcomings of alignment time-consuming and low precision in the coaxial debugging of existing optical fiber rod deposition machine tools, a set of automatic positioning and alignment device with high accuracy and high stability was designed and implemented. First, the lens distortion and other errors were eliminated by the Zhang’s calibration method. Then, in order to eliminate the influence of noise and background on the subsequent processing, the images were pre-processed using Gaussian filtering, dynamic threshold and canny edge detection to extract accurate image edges. Finally, a new arc classification circular detection algorithm was used to achieve the precise positioning of the center of the measured surface in the image the scale and deviation of the image were determined according to the fitted circle and the position of the machine tool chuck was adjusted to achieve the center alignment of both ends of the machine tool. Comparing the algorithm in this paper with several common circle detection methods, the average error and average positioning time of our algorithm are 1.0664 pixel and 244 ms, respectively, indicating that our algorithm is much better than other detection algorithms. The experimental results show that the accuracy and efficiency of the device for machine tool alignment are significantly improved. When the camera resolution is 500 w, the system detection accuracy is less than 0.1 mm. Our work meets the requirements of machine tool coaxial inspection, and provides a fast and effective method for machine tool coaxial debugging

常用功能