The Study Of Portable Fiber End Face Inspection

With the development of optical fiber communication, it is great significance to detect the fiber end conveniently for network connections. In this paper, we designed a portable optical fiber end-detection equipment. It is small and easy to operate in the outdoors. We set up the fiber end face cleanliness standards according to the U.S. Connecting Electronics Industries standards IPC-8497-1 Cleaning Methods and Contamination Assessment for Optical Assembly. The images were collected by the ARM processor for processing at real time and determined cleanliness of the fiber end face according to the standards. The system has a good human-computer interaction, good storage model, and fully functions.Major work of this dissertation:1. The entire system was designed. It includes optical system, lighting systems, image acquisition system and hardware platform choices.2. The image acquisition and display process of the fiber end face detection equipment on the ARM processor was achieved by invoking the bottom driver. The human-computer interaction module was designed to adjust the threshold of the parameter, and the adaptability of the software was enhanced.3. The whole image processing program of the equipment was designed. The Canny edge detection algorithm was designed to detect the edge of good quality and continuous. An improved Hough algorithm was proposed to detect the position of the fiber cladding. Binary process was used regional separation according to the light distribution of the fiber end image. The stain and the scratch on the fiber end image was distinguished according to the definition of roundness. The size of the stain's the smallest external circle and the scratch's length was figured out in order to judge the quality of optical fiber end face cleaning.4. The mechanical structure of the fiber end face detection equipment was improved. Its outstanding characteristics are small, simple structure, and easy to operate outfield.5. The entire system was assembled, tested and experimented. The experimental results show that: In the improved algorithm, the accuracy of the fiber end face detection equipment was 92%, and processing time is about 6s. It basically reached the requirements of real-time processing.