Rapid detection and identification of fiber-doped by laser-induced breakdown spectroscopy and machine learning

The rapid detection and identification of fibers is of great significance for the recycling of fiber in the experiment. However, the analysis of these fibers is extremely challenging due to extreme similarities in appearance or chemical composition. In this study, a system based on laser induced breakdown spectroscopy (LIBS) and machine learning algorithms is used for identifying and classifying fibers tested. Three different kinds of fibers are selected as samples for detection. Principal component analysis (PCA) is used to distinguish the three fibers and training data set is stored for further identification. The result show that these fibers recognition rate reaches 84.5%. The results further verify that LIBS combined with machine learning could be adopted to detect fiber-doped elements, which provides a new method for fiber detection and classification.