Bayesian inference approach for Full Poincaré Mueller polarimetry

Full Poincaré Mueller Polarimetry is a new technique for characterizing samples by means of their Mueller matrix. The method is based on the use of a full Poincaré beam as a generator of polarization states. These beams present different polarization states, covering the entire Poincaré sphere surface, at different points in the beam cross section. To obtain the Mueller matrix, Stokes parameters are collected at both the entrance and the output of the sample. They are calculated from irradiance measurements at each pixel of a CCD camera for different configurations of the polarization state analyzer. These measurements can be processed in several ways. In this work, we propose to use Bayesian inference, in particular, Markov chain Monte Carlo methods, to obtain, without any prior knowledge of the sample, its Mueller matrix together with its uncertainties. The new approach is tested with experimental measurements of different samples and compared with the real theoretical Mueller matrices. Excellent agreement is observed between the experimental results and the theoretical ones for all the samples tested. •A Bayesian approach is used to analyze full Poincaré Mueller polarimetry measurements.•All Mueller matrix elements are characterized by their probability density function.•Flexibility in specifying suitable models for the data.•It generates robust estimates of the parameters and their credible intervals.