-
摘要:
无论是通讯光纤还是医用光纤,光纤的几何参数总是评价其质量的重要指标。灰度法是国标GB15972.20-2008中的建议方法,但该方法在拟合过程中会出现拟合圆与椭圆的中心不重合,存在测量原理上的缺陷。且当光纤的切割效果与照明条件发生改变时,往往导致测量数据的不稳定并带来误差。本文用更符合光纤端面实际的任意椭圆函数(非标准椭圆),且仅用这一种函数拟合的方法求取光纤几何参数,从而从根本上消除由圆拟合与椭圆拟合的中心不一致带来的原理缺陷。同时,由于在计算各个参数时不需要图像分布灰度的具体值,从而降低了对测量条件的要求。实验表明,本文方法能有效提高仪器测量结果的稳定性和一致性。
Abstract:The fiber geometry of communication fibers and medical fibers are always standards to evaluate the quality of optical fibers. The measurement of fiber geometry with gray scale method is one of the commonly used measurement methods. It is also the proposed method in the national standard GB15972.20-2008. In this method, the fiber geometry is obtained by fitting the elliptical curve and fitting the circular curve in two steps, but the center of the two curves may not be coincided. Thus, there is a defect in the measurement principle in the method. The measurement of fiber geometry with gray scale method has a high requirement for cutting effects and lighting conditions. When measurement conditions change, it often leads to the instability of the measured data and brings errors. In this paper, we use the arbitrary elliptical function (non-standard ellipse) which is more suitable for the fiber end face, and only use this function fitting method to get the fiber geometry to fundamentally eliminate the principle defect caused by the inconsistent center fitting between the circle fitting and the ellipse fitting. At the same time, the requirement of measurement condition is reduced, because the specific value of image distribution grayscale is not needed when calculating each parameter. Experiments show that this method can effectively improve the stability and consistency of measurement results.
-
Key words:
- optical fiber geometry /
- gray scale method /
- ellipse curve fitting
-
Overview: The fiber geometry of communication fibers and medical fibers are always important parameters to evaluate the quality of optical fibers. The fiber geometry mainly includes the diameter and ovality of claddings and cores, and the concentricity error of claddings and cores. The measurement of fiber geometry with gray scale method is one of the commonly used measurement methods proposed in the national standard GB15972.20-2008. In the gray scale method, the fiber geometry is obtained by two-step fittings that are the fitting of circular curve and the fitting of ellipse curve. However, the geometric centers of the two fitting curves may not necessarily coincide, causing the measurement inaccuracy of fiber geometry. Obviously, there is a defect of the measurement principle in the method. The measurement of fiber geometry with gray scale method has a high requirement for cutting effects and lighting conditions. When measurement conditions change, it often leads to the instability of the measured data and brings errors. This paper proposes a method for obtaining the optical fiber geometry with fitting the edge of optical fiber in an arbitrary ellipse curve (non-standard ellipse) which more coincide the real optical fiber end face.
The method is mainly divided into three steps: image preprocessing, edge extraction and ellipse curves fitting. The first step, image preprocessing, is to eliminate some of the noise errors in the image by median filtering the image, in order to make the subsequent edge extraction better. The second step, edge extraction, is to use the Canny operator to extract the image of the optical fiber end face. At this time, there are still some noise signals and false edges at the edge of the optical fiber. The third step, ellipse curves fitting, is to fit the edge data points with the arbitrary ellipse, and to set an appropriate threshold value at the edge of the fitting ellipse curve, then is to remove the data points beyond the threshold value as error data points. All optical fiber geometry can be calculated by arbitrary ellipse curve in one-step fitting, so eliminating the measurement error caused by the center inconsistency between the circle curve fitting and the ellipse curve fitting. At the same time, because the specific value of the image distribution gray scale is not required when calculating each parameter, the edge data of the optical fiber is used for fitting, thereby the measurement condition is effectively reduced. Taking fiber core data as an example, the data of diameter and ovality measured by the standard instrument are 8.420 μm and 0.670%, respectively. When cutting effect of fiber end face or lighting condition is poor, the instrument data become 9.436 μm and 2.016%, while the data measured in this paper are 8.804 μm and 0.553%, respectively. Experiment results show that the method can significantly improve the accuracy and precision of the measurement results of instruments.
-
表 1 FGM-5几何参数测试仪的正常测量数据
Table 1. Standard data of FGM-5 optical fiber geometry measurement system
实验组数 包层直径/μm 包层不圆度/% 纤芯直径/μm 纤芯不圆度/% 芯-包同心度/μm 1 125.073 0.116 8.421 0.865 0.241 2 125.094 0.117 8.409 0.335 0.236 3 125.120 0.118 8.431 0.666 0.246 4 125.101 0.098 8.449 0.616 0.247 5 125.093 0.102 8.388 0.868 0.221 平均值 125.096 0.110 8.420 0.670 0.238 最大偏差 0.024 0.012 0.032 0.335 0.017 表 2 实测的几何参数数据
Table 2. Measured fiber geometry data
实验组数 包层直径/μm 包层不圆度/% 纤芯直径/μm 纤芯不圆度/% 芯-包同心度/μm 1 125.081 0.108 8.589 0.465 0.210 2 125.101 0.114 8.551 0.208 0.191 3 125.125 0.111 8.575 0.751 0.198 4 125.108 0.122 8.583 0.862 0.188 5 125.097 0.114 8.532 0.446 0.201 平均值 125.102 0.114 8.566 0.546 0.198 最大偏差 0.023 0.008 0.034 0.338 0.012 表 3 FGM-5几何参数测试仪的非正常测量数据
Table 3. Deviant data of FGM-5 optical fiber geometry measurement system
实验组数 包层直径/μm 包层不圆度/% 纤芯直径/μm 纤芯不圆度/% 芯-包同心度/μm 1 124.832 0.077 9.092 1.529 0.288 2 124.830 0.094 9.037 1.148 0.302 3 124.869 0.089 9.037 1.754 0.302 4 124.862 0.095 9.045 0.837 0.283 5 124.837 0.091 9.010 2.016 0.278 平均值 124.846 0.089 9.044 1.457 0.291 最大偏差 0.023 0.012 0.048 0.62 0.013 表 4 实测的几何参数数据
Table 4. Measured fiber geometry data
实验组数 包层直径/μm 包层不圆度/% 纤芯直径/μm 纤芯不圆度/% 芯-包同心度/μm 1 124.874 0.084 8.638 0.654 0.245 2 124.857 0.072 8.643 0.312 0.252 3 124.897 0.069 8.701 0.635 0.234 4 124.885 0.078 8.701 0.508 0.252 5 124.862 0.055 8.638 0.357 0.232 平均值 124.875 0.072 8.664 0.493 0.243 最大偏差 0.022 0.017 0.037 0.181 0.011 表 5 光纤几何参数均值数据对比
Table 5. Comparison of mean data of optical fiber geometry
实验方法 包层直径/μm 包层不圆度/% 纤芯直径/μm 纤芯不圆度/% 芯一包同心度/μm FGM-5数据 正常端面 125.096 0.110 8.420 0.670 0.238 图 4(a) 124.846 0.089 9.044 1.457 0.291 图 4(b) 125.136 0.220 9.436 2.016 0.306 与正常端面测量值偏差 图 4(a) 0.250 0.021 0.624 0.787 0.053 图 4(b) 0.040 0.110 1.016 1.346 0.068 任意椭圆函数拟合法 正常端面 125.102 0.114 8.566 0.546 0.198 图 4(a) 124.875 0.072 8.664 0.493 0.243 图 4(b) 125.104 0.052 8.804 0.553 0.270 与正常端面测量值偏差 图 4a) 0.227 0.042 0.098 0.053 0.045 图 4(b) 0.002 0.062 0.238 0.007 0.072 -
[1] 中华人民共和国国家质量监督检验检疫总局, 中国国家标准化管理委员会.光纤试验方法规范第20部分: 尺寸参数的测量方法和试验程序光纤几何参数: GB 15972.20–2008[S].北京: 中国标准出版社, 2008.
General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China, Standardization Administration of the People's Republic of China. Specifications for optical fibre test methods-part 20: measurement methods and test procedures for dimensions- fiber geometry: GB 15972.20–2008[S]. Beijing: Standards Press of China, 2008.
[2] 陈磊, 陈进榜, 陆润华.光纤几何参数的自动检测仪[J].光学学报, 2001, 21(10): 1245–1248. doi: 10.3321/j.issn:0253-2239.2001.10.021
Chen L, Chen J B, Lu R H. Automatic measurement of optical fiber geometric parameters[J]. Acta Optica Sinica, 2001, 21(10): 1245–1248. doi: 10.3321/j.issn:0253-2239.2001.10.021
[3] 殷爱娥, 姜仲玄, 张一龙.光纤折射率剖面的折射近场法测量的研究[J].光学学报, 1989, 9(2): 181–185. doi: 10.3321/j.issn:0253-2239.1989.02.014
Yin A E, Jiang Z X, Zhang Y L. Refracted near-field technique for the measurement of optical fiber refractive index profiles[J]. Acta Optica Sinica, 1989, 9(2): 181–185. doi: 10.3321/j.issn:0253-2239.1989.02.014
[4] 高迎春.基于折射近场法测量光纤折射率分布的仿真研究[D].哈尔滨: 哈尔滨工程大学, 2012.
Gao Y C. Simulation of measuring the optical fiber refractive index profiles by refraction near-field method[D]. Harbin: Harbin Engineering University, 2012.
[5] 孙会刚, 储九荣, 钟力生, 等.塑料光纤折射率分布的测量方法[J].光纤与电缆及其应用技术, 2001(4): 12–16. doi: 10.3969/j.issn.1006-1908.2001.04.004
Sun H G, Chu J R, Zhong L S, et al. Measurement of refractive-index profile of plastic optical fibers[J]. Optical Fiber & Electric Cable, 2001(4): 12–16. doi: 10.3969/j.issn.1006-1908.2001.04.004
[6] 贾宏志, 李育林, 忽满利.光纤光栅的制作方法[J].激光技术, 2001, 25(1): 23–26. doi: 10.3969/j.issn.1001-3806.2001.01.001
Jia H Z, Li Y L, Hu M L. Fabrication methods of fiber gratings[J]. Laser Technology, 2001, 25(1): 23–26. doi: 10.3969/j.issn.1001-3806.2001.01.001
[7] 武志宏.基于CCD和图像处理技术的六角光纤几何尺寸测量系统的研究[D].太原: 太原理工大学, 2006.
Wu Z H. The research about measurement system of six-angle fiber geometrical dimension on the basis of CCD and image processing[D]. Taiyuan: Taiyuan University of Technology, 2006.
[8] 马利东.光子晶体光纤有效模面积测量技术的研究[D].秦皇岛: 燕山大学, 2016.
Ma L D. Research on effective mode area measurement of photonic crystal fiber[D]. Qinhuangdao: Yanshan University, 2016.
[9] Hameed M F O, Obayya S S A. Modal analysis of a novel soft glass photonic crystal fiber with liquid crystal core[J]. Journal of Lightwave Technology, 2012, 30(1): 96–102. doi: 10.1109/JLT.2011.2175436
[10] Rosa L, Coscelli E, Poli F, et al. Thermal modeling of gain competition in Yb-doped large-mode-area photonic-crystal fiber amplifier[J]. Optics Express, 2015, 23(14): 18638–18644. doi: 10.1364/OE.23.018638
[11] 杨远.基于机器视觉的光纤几何参数检测研究[D].哈尔滨: 哈尔滨工程大学, 2011.
Yang Y. Research on detecting optical fiber geometric parameter based on machine vision[D]. Harbin: Harbin Engineering University, 2011.
[12] 凌凤彩, 康牧, 林晓.改进的Canny边缘检测算法[J].计算机科学, 2016, 43(8): 309–312. doi: 10.11896/j.issn.1002-137X.2016.8.063
Ling F C, Kang M, Lin X. Improved Canny edge detection algorithm[J]. Computer Science, 2016, 43(8): 309–312. doi: 10.11896/j.issn.1002-137X.2016.8.063
[13] 郭萌, 胡辽林, 赵江涛.基于Kirsch和Canny算子的陶瓷碗表面缺陷检测方法[J].光学学报, 2016, 36(9): 0904001. doi: 10.3788/aos201636.0904001
Guo M, Hu L L, Zhao J T. Surface defect detection method of ceramic bowl based on Kirsch and Canny operator[J]. Acta Optica Sinica, 2016, 36(9): 0904001. doi: 10.3788/aos201636.0904001
[14] 王文豪, 姜明新, 赵文东.基于Canny算子改进的边缘检测算法[J].中国科技论文, 2017, 12(8): 910–915. doi: 10.3969/j.issn.2095-2783.2017.08.013
Wang W H, Jiang M X, Zhao W D. Edge detection algorithm based on improved Canny operator[J]. China Sciencepaper, 2017, 12(8): 910–915. doi: 10.3969/j.issn.2095-2783.2017.08.013
[15] 王万国, 王仕荣, 徐正飞, 等.基于边界的最小二乘椭圆拟合改进算法[J].计算机技术与发展, 2013, 23(4): 67–70. doi: 10.3969/j.issn.1673-629X.2013.04.016
Wang W G, Wang S R, Xu Z F, et al. Optimal ellipse fitting algorithm of least square principle based on boundary[J]. Computer Technology and Development, 2013, 23(4): 67–70. doi: 10.3969/j.issn.1673-629X.2013.04.016
[16] 熊风光, 李希, 韩燮.基于整体最小二乘的椭圆拟合方法[J].微电子学与计算机, 2017, 34(1): 102–105. http://d.old.wanfangdata.com.cn/Periodical/wdzxyjsj201701022
Xiong F G, Li X, Han X. A method of ellipse fitting based on total least squares[J]. Microelectronics & Computer, 2017, 34(1): 102–105. http://d.old.wanfangdata.com.cn/Periodical/wdzxyjsj201701022
[17] 曹俊丽, 李居峰.基于莱特准则的椭圆拟合优化算法[J].计算机应用, 2017, 37(1): 273–277. doi: 10.11772/j.issn.1001-9081.2017.01.0273
Cao J L, Li J F. Improved ellipse fitting algorithm based on Letts criterion[J]. Journal of Computer Applications, 2017, 37(1): 273–277. doi: 10.11772/j.issn.1001-9081.2017.01.0273
[18] Gander W, Golub G H, Strebel R. Least–squares fitting of circles and ellipses[J]. BIT Numerical Mathematics, 1994, 34(4): 558–578. doi: 10.1007/BF01934268
[19] Mitchell D R G, Van den Berg J A. Development of an ellipse fitting method with which to analyse selected area electron diffraction patterns[J]. Ultramicroscopy, 2016, 160: 140–145. doi: 10.1016/j.ultramic.2015.10.009
[20] 何晨程, 杨婧, 宋海燕, 等.切割刀老化对单模光纤几何尺寸检测结果的影响[J].现代传输, 2016(1): 56–61. doi: 10.3969/j.issn.1673-5137.2016.01.001
He C C, Yang J, Song H Y, et al. Effect of cutting tool aging on the geometric size measurement of single mode fiber[J]. Modern Transmission, 2016(1): 56–61. doi: 10.3969/j.issn.1673-5137.2016.01.001
[21] 刘为, 唐春晖, 马秀梅, 等.缺陷光纤端面几何参数的测量[J].光通信研究, 2013, 39(6): 35–38. doi: 10.3969/j.issn.1005-8788.2013.06.011
Liu W, Tang C H, Ma X M, et al. Measurement of geometric parameters of defective fiber ends[J]. Study on Optical Communications, 2013, 39(6): 35–38. doi: 10.3969/j.issn.1005-8788.2013.06.011