Research On Geometric Parameter Measurement And Defect Detection Method Of Single-mode Optical Fiber

The geometrical parameters of single-mode optical fiber not only affect the optical transmission and mechanical properties of the optical fiber,but also play an important role in the splice loss at the junction.The defects in the single-mode optical fiber will not only affect the quality of the optical fiber itself,but also do the damage to the whole system.So it is very important to measure geometric parameters and detect defects of single-mode optical fiber accurately.At present,the geometric parameters and defects of single-mode fiber are mainly detected by manual method,which has the disadvantages of low efficiency and large error.In this paper,based on the image features of single-mode optical fiber,the image information of single-mode optical fiber is acquired by machine vision technology,and the image information is processed and analyzed.The paper is divided into three parts.Firstly,a single-mode optical fiber's end-face image acquisition system is designed based on machine vision technology.The obtained fiber's end-face image is processed by image enhancement and filtering,and then Canny operator is used for edge feature extraction.The edge feature is extracted,and the edge feature information is fitted by the cubic spline function of circle to get the geometric parameters of single-mode optical fiber.At the same time,camera calibration and standard optical fiber calibration are carried out to ensure the accuracy of the whole measurement system.The experimental results show that the repeatability of the method is better than that of the imported equipment,and the repeatability of each parameter test is improved by 20%.Secondly,according to the main factors that affect the detection accuracy in single-mode optical fiber geometry detection,an autofocus and fiber end-face quality detection method is proposed.The autofocus and optical fiber's end-face quality detection are mainly to ensure that the obtained end-face image is clear and visible,and to reduce the error of the image processing.The autofocus method proposed in this paper is mainly based on the depth-of-focus method,choosing the Tenengrad function with higher focusing accuracy as the focusing evaluation function,and then using the hill-climbing method to carry out the focusing search,which can carry out real-time focusing.In the optical fiber's end-face quality detection,a method based on focal plane is proposed to measure the position of the clear points in different angles,the difference between this position and the clear position of the whole end-face imaging is taken as the evaluation value of the end-face quality of the optical fiber.In the detection of optical fiber's end-face quality,a method based on the focal plane is proposed to measure the position of the image's clear points at different angles,the difference between this position and the clear position of the whole end-face imaging is taken as the evaluation value of the end-face quality of the optical fiber.Finally,based on machine vision technology,a method of optical fiber defect recognition and judgment is proposed.The method is mainly used to obtain side images of optical fiber at different angles by machine vision system,including coating layer.Then the image is filtered,and the edge information is extracted,which is divided into regions according to the edge feature information to realize the automatic recognition and judgment of defect types.The experimental results show that the accuracy of defect detection can reach 100%.