Polarimetric Imaging
Overview
Polarimetric imaging captures how light’s polarization changes through reflection, scattering, or transmission, often represented with Stokes or Mueller parameters. Researchers use polarization cameras and computational reconstruction for tasks such as image demosaicking and restoration, surface-normal and material estimation, biomedical imaging, and remote-sensing classification with polarimetric synthetic-aperture radar (PolSAR). Common approaches combine optical-physics constraints with deep networks, including diffusion models, complex-valued architectures, and learned priors; evaluation spans reconstruction benchmarks, shape and classification datasets, and real-world capture settings. A central challenge is balancing physical consistency and robustness against limited or mismatched training data, especially when measurements are noisy or some polarization quantities are not identifiable from the available images.
Latest papers
35 papers5 papers in the last four weeks, up 67% on the four weeks before. 0.0% of all new papers.