| Popis: |
For different kinds of applications, mainly focused on probing/imaging of complex media, management of light scattering remains a challenge. Numerous studies have proven that the statistical laws followed by the intensity patterns and the polarimetric behaviour of the scattered field at the speckle scale provide a key solution to discriminate between surface and bulk scattering. It was shown for totally diffusing media that surface and bulk scattering can be seen as extreme scattering regimes. For these two regimes, exact electromagnetic calculation can be performed, but this requires a deterministic knowledge of the medium (surface profile, inhomogeneity function) under study. Furthermore these exact models are time and memory consuming, which reduces the analysis to samples with small dimensions. Within this framework statistical optics provide solutions which allow to simplify the problem. As an illustration, in previous studies our group recently proposed to use random phasors matrices to predict the statistical behaviour of surface and bulk scattering patterns in terms of intensity and polarimetric parameters. These phenomenological results were validated by comparison to both exact electromagnetic theories and experimental data, with high agreement. However in these works only the extreme scattering regimes were addressed while they involve zero or total correlation of the scattering coefficients. In this paper we show how to extend the analysis to intermediary scattering regimes, with a comparison to experiment. Furthermore we propose a simplified model involving a unique correlation parameter in the scattering matrix of a strongly disordered medium. Comparison with experiment emphasizes the validity of this model to predict the statistical behavior of the speckle patterns and their polarization properties, as well as spatial depolarization and temporal repolarization. The parameter is connected with the weight of multiple scattering and allows to consider transition between surface and bulk scattering. This statistical approach provides great help to analyze scattering media in the absence of electromagnetic theories. |
