Polarization state classification of vector vortex beams using the PaCMAP algorithm

Vector vortex beams (VVBs), characterized by their anisotropic wavefront and polarization distributions, offer a multitude of optical field degrees of freedom. These properties endow VVBs with significant potential for applications in optical communication, quantum technology, high-resolution imaging, optical tweezers, and other fields. Accurate identification of the modes and polarization distributions of VVBs is crucial for their successful application. However, the performance of existing principal component analysis (PCA) methods for classifying the states of VVBs deteriorates markedly in disturbed environments. To address this challenge, this paper proposes a joint algorithm based on pairwise controlled manifold approximation projection (PaCMAP) for classifying and recognizing VVBs in complex noise environments. The classification performance of this algorithm on the states of VVBs is experimentally verified. Results indicate that the PaCMAP-based joint algorithm significantly enhances the classification of VVBs' polarization states, thereby providing a reliable method for the development and application of vector vortex light.