Sun Minghao, He Honghui, Zeng Nan, Du E, Guo Yihong, Peng Cheng, He Yonghong, Ma Hui
Appl Opt. 2014 May 10;53(14):2949-55. doi: 10.1364/AO.53.002949.
Polarization parameters contain rich information on the micro- and macro-structure of scattering media. However, many of these parameters are sensitive to the spatial orientation of anisotropic media, and may not effectively reveal the microstructural information. In this paper, we take polarization images of different textile samples at different azimuth angles. The results demonstrate that the rotation insensitive polarization parameters from rotating linear polarization imaging and Mueller matrix transformation methods can be used to distinguish the characteristic features of different textile samples. Further examinations using both experiments and Monte Carlo simulations reveal that the residue rotation dependence in these polarization parameters is due to the oblique incidence illumination. This study shows that such rotation independent parameters are potentially capable of quantitatively classifying anisotropic samples, such as textiles or biological tissues.
偏振参数包含了关于散射介质微观和宏观结构的丰富信息。然而，其中许多参数对各向异性介质的空间取向敏感，可能无法有效揭示微观结构信息。在本文中，我们在不同方位角下对不同纺织品样本进行了偏振成像。结果表明，通过旋转线性偏振成像和穆勒矩阵变换方法得到的对旋转不敏感的偏振参数可用于区分不同纺织品样本的特征。进一步通过实验和蒙特卡罗模拟研究发现，这些偏振参数中残留的旋转依赖性是由斜入射照明引起的。该研究表明，此类与旋转无关的参数有可能对各向异性样本（如纺织品或生物组织）进行定量分类。
