Accurate detection technology for polarization axis of Panda polarization-maintaining fiber

The detection of polarization axis is a key technology in application of polarization-maintaining fiber(PMF). In the azimuth system for PMF based on image processing, the false edge of the stress circle in edge detection made the center location incorrect. In order to solve the problem, a method based on least square fitting was proposed. This method wiped out the incorrect edge in the process of iteration. When the number of point for iteration was smaller than a threshold, it came back to the original set of point for iteration. Experimental result show that the algorithm can avoid the disturbance made by noise and the incorrect edge. When coming to the real cross-section image, it shows strong robustness.