Machine Vision Based Internal Defects Inspection For Optical Fiber Preform

In recent years, with the popularization of information technology, the optical fiber is terribly needed which lead to more and more companies begin to enter the fiber optic industry. For the manufacture of optical fiber, optical fiber preform is one of the most important materials since its quality determines the quality of optical fiber. If there are impurities and air bubbles inside perform, then the optical fiber would be easily broken when being drawn. Thus, internal defects inspection for optical fiber perform is very important in the manufacture of optical fiber.Currently, the artificial detection for internal defect detection of optical fiber perform is mainly adopted, which has the disadvantage of high error detection rate, low efficiency and lack of unified standard to judge internal defects. In view of the shortcoming of artificial detection, this paper put forward machine vision based detecting method for internal defects of the optical fiber preform. The advantages of this method are obviously:high processing speed, standarded and unified testing, easy to control, and the online automatic detection can be realized. The purpose of this article is to design and develop an online automatic detection system for internal defects inspection of optical fiber preform according to detecting requirement. The main research content is summarized as follows:The thesis introduced the important role of the optical fiber preform in fiber manufacturing process and the current situation of fiber industry and fiber preform production status in China. The basic theory of machine vision and its development and application of both at home and abroad, the open source computer vision library (OpenCV) commonly used in the development of machine vision are introduced.According to the testing requirements of optical fiber preform, the thesis designed and developed a set of internal defects detection system, including light source system, image acquisition system, motion control system and image processing system. Each part has been introduced in detail. The parameters need to be considered when choosing light source, camera, video camera parameters, and the devices we adopted in the system are presented in the paper. In this paper, the defects recognition and extraction algorithm, including image denoising, image segmentation, image binarization, defect extraction, defect position description and defect characteristics is proposed based on Open CV. We designed a set of system software by using VS2008based on MFC in the Windows xp platform, including motion control module with the Windows xp platform, camera control module, image processing module and results show that the module. Each module has been introduced in detail.The system developed in this paper has been used to measure the internal defects of the optical fiber preform, and the factors affected the accuracy of detection also have been analyzed. The experimental results show that the system can effectively detect the internal defect of the optical fiber perform. The research results demonstrated in this paper also can be applied to the detection of internal defects in other optical materials, which has a certain use value.