The polarization devices based on polarization-maintaining fibers are extremely valuable in the fields of coherent fiber-optic communications,fiber-optic gyroscopes,and fiber-optic sensors.The quarter-wave plate is a polarization device that can convert linearly polarized light into circularly polarized light and has very important application value in full optical fiber current sensor.The quarter-wave plate is usually made by welding two pieces of polarization-maintaining fiber at 45° on the axis,and then cutting the quarter-beat length of the wavelength plate fiber.The axial angle control of the polarization-maintaining fiber is a key step in the manufacturing process,which directly affects the performance of the quarter-wave plate.How to precisely control the axial angle of the polarization-maintaining fiber is not only a difficult point in the field of quarter-wave plate fabrication but also becomes the biggest bottleneck in the development of quarter-wave plates.In view of the above problems,this paper proposes a polarization-maintaining fiber alignment system based on the Xilinx ZYNQ platform.This system,which adopts ARM+FPGA architecture,is divided into system hardware platform and software processing platform.The hardware platform of the system has designed and integrated the modules including image acquisition,median filtering,image display and motor control by utilizing the high computing efficiency and high flexibility of FPGA.The image acquisition module realizes 2 channels of optical fiber image acquisition with a resolution of 640×480;the median filter module realizes real-time filtering processing of the acquired images,effectively filtering the noise in the original image and improving the imaging quality;the image display module realizes the splicing and display of two optical fiber images,the overall display effect is stable,and there is no image tearing and dislocation.The software processing platform is based on the ARM.After reading the optical fiber image data collected by the system hardware platform,the research on the image recognition algorithm is carried out by using the high scalability of the ARM.There are three modules including optical fiber end face angle measurement,optical fiber displacement control,and optical fiber rotation control,and theoretical analysis and experimental tests are carried out on the module designs.The results show that the optical fiber end face angle measurement module can accurately identify the optical fiber end face angle angle within1° and meet the requirement of polarization-maintaining welding.The optical fiber displacement and optical fiber rotation control module realize the 45° axis alignment function,and the axis identification accuracy is in the range of 1.1° to 1.4°,which is basically at the same level as the current mainstream polarization maintaining fiber axis identification technology.The system hardware platform and software processing platform constitute a complete polarization-maintaining fiber alignment system,which is used to complete the preparation of quarter-wave plates to verify the actual operating performance of the system.In this paper,the system is used to complete the manufacture of the quarter-wave plate,and the experimental detection and performance analysis are carried out.The experimental result shows that the extinction ratio of the resulting quarter-wave plate is less than 1 d B,and the degree of polarization is higher than 90%.The performance of the quarter-wave plate is fine. |