Creating a Polarization Image


In the backscattering of light, a two dimensional image representing the distribution of intensity of scattered light can be acquired using a CCD camera. Each point in the image directly corresponds to one point on the surface of the scattering medium. If the medium was represented by one matrix of 16 numerical elements, every possible point in the medium would effect the polarization of the laser light in the exact same way. We know this is not true. The effect that a given point of the substance has on light depends on how many times the photons were re-radiated by particles before escaping the medium altogether and being collected into the CCD element. Therefore a different matrix must be created for every possible part of the medium. The media used in backscattering experimentation is generally arranged to have a surface area greater then one square centimeter. In this area, slightly smaller than a thumbnail, thousands of individual 4x4 matrices would be needed for the thousands of possible paths the photons from the laser could take.

The beauty of images is their ability to represent a huge amount of numerical data in a clean, straightforward manner. Every point in the scattering medium is represented by 16 numerical values in matrix form. All possible numerical values can be translated to a corresponding color value. Thus the matrix becomes 16 dots of vivid color. But when the thousands of matrices of colored dots are combined to represent the entire surface area of the medium, the dots join to form spatial arrays of color value: pictures. These pictures are easy to interpret -- distinguishing differing optical properties in a substance is as easy as telling red from blue.





[Title Page] [The Mueller Matrix] [Optical Arrangements]