In earlier work we introduced a metric based on the spatial variability of the Angle of Polarization (AoP) as a measure of the quality of the polarization signature measured by a polarimeter. In this paper, we call this the spatial angle of polarization confidence metric, or S-AoPC. The S-AoPC is computed by considering the variability of the AoP in a small region surrounding each pixel. It is based on the assumption that high spatial variance in AoP exists when the underlying Stokes parameters are essentially random. In the present paper we extend this concept to also consider temporal variation of AoP in a time angle of polarization confidence metric, or T-AoPC. We examine the S-AoPC and T-AoPC empirically for a range of polarization images and we consider applications of these metrics in polarimetry including polarimetric display, setting the parameters of the camera such as integration time, and even helping to understand the processing of hybrid polarimeter data to maximize image quality. As the name implies, the angle of polarization confidence metric can be extended to consider variability in other dimensions such as wavelength for different types of polarimeters.
|