光强极值法测量锥型光纤半径的理论与技术研究

发布时间：2018-01-16 14:28

本文关键词：光强极值法测量锥型光纤半径的理论与技术研究　出处：《西安电子科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：锥型光纤作为一种特殊的光纤,普遍应用在医学、近场光学显微镜以及光孤子通信等领域。而对于锥型光纤半径的准确测量,是研究锥型光纤特性以及提高其生产质量的前提和保证。本文在国内外光纤半径测量技术研究的基础上,重点研究基于光散射原理的光强极值法,并对其测量方案进行改进,以提高测量的速度和精确度。该方案将测量过程分为半径数据库建立、图像处理和数据匹配三大部分进行研究。(1)根据散射原理公式建立相应数据库：从基本的波动方程出发,分析了高斯光束透过均匀透明的锥型光纤介质后透射、衍射光的复杂叠加过程,并引入相关函数对散射公式进行近似求解。然后,通过原理公式,在给定的半径取值范围内,获取每一半径值所对应的散射光强值与散射角关系图谱,并从图谱中提取每一个光纤半径的一套散射角数据,建立数据库。(2)搭建实验平台获取散射图像并进行相关图像处理：本文实验中所使用的光纤采用了中国航空自动控制第618研究所的单模锥型光纤,对其锥区半径进行了测量实验。测量系统包括：激光光源整形系统、柱面透镜望远系统以及图像采集系统。其中,对于激光光源束腰半径的透镜变换过程、柱面透镜望远系统的参数选择等方面进行了详细的计算分析。之后,由工业相机拍摄得到散射条纹图像,对图像进行了以下方案的预处理：先转换图像格式,再截取图像,分别采用直方图均衡化、均值滤波、低通滤波和中值滤波四种平滑滤波方法处理后,通过比较,选择了先直方图均衡化,再使用中值滤波进行图像处理。将处理过的散射图像中的单列光强极值图像用平滑滤波算法进行平滑整形,并从中提取明暗条纹光强极值所对应的散射角数据。(3)设计了数据匹配的方案,该方案的主要依据是最小二乘理论。将图像处理中获得的光强极值散射角数据与自建数据库的相应散射角数据进行最优匹配,获取锥区半径值,并输出测量结果。经过实验测试,锥型光纤半径测量系统可以准确、快速的测量锥型光纤的锥区半径,并对实际测量精度及性能差异作了评估研究,提出了进一步提高系统测量性能的改进方向,发现了算法上的不足之处,提出了改进的措施。
[Abstract]:As a special fiber conical fiber is widely used in medicine near-field optical microscope and optical soliton communication and so on. It is the premise and guarantee to study the characteristics of conical fiber and improve its production quality. Based on the research of optical fiber radius measurement technology at home and abroad, the light intensity extreme value method based on the principle of light scattering is mainly studied in this paper. The measurement scheme is improved to improve the speed and accuracy of the measurement, which divides the measurement process into radius database. Three parts of image processing and data matching are studied. (1) according to the scattering principle formula, the corresponding database is established: starting from the basic wave equation. The complex superposition process of diffracted light and transmission of Gao Si beam through a uniform and transparent tapered fiber medium is analyzed, and the scattering formula is solved approximately by introducing the correlation function. Then, the principle formula is adopted. In the range of given radius, the relationship between the scattering intensity and the scattering angle of each radius is obtained, and a set of scattering angle data of each optical fiber radius is extracted from the spectrum. Build the experimental platform to obtain scattering images and image processing: the optical fiber used in this experiment uses the single-mode conical fiber of China Aviation automatic Control Institute 618. The measurement system includes laser light source shaping system, cylindrical lens telescope system and image acquisition system. The parameter selection of cylindrical lens distant system is calculated and analyzed in detail. Then the scattered fringe image is captured by industrial camera and the image is preprocessed as follows: first transform the image format. After intercepting the image, the histogram equalization, mean filter, low-pass filter and median filter are used to process the image. Through comparison, the first histogram equalization is selected. Then the median filter is used to process the image, and the single-row intensity extremum image of the scattered image is smoothed by the smoothing filtering algorithm. The scattering angle data corresponding to the extreme value of light intensity of dark and light stripes are extracted from the data. 3) the scheme of data matching is designed. The main basis of the scheme is the least square theory. The scattering angle data obtained in image processing and the corresponding scattering angle data of the self-built database are matched optimally to obtain the radius of the cone region. The experimental results show that the tapered fiber radius measurement system can accurately and quickly measure the cone radius of the tapered fiber, and the actual measurement accuracy and performance difference are evaluated and studied. The improvement direction of further improving the measurement performance of the system is put forward, the shortcomings of the algorithm are found, and the improvement measures are put forward.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN253;TP391.41

【相似文献】