Strict registration is critical for imaging polarimetry; a highly accurate registration approach to the coronal image and imaging polarimetry of the Fe 6374 angstrom red coronal line have been researched in this paper. In order to improve the registration accuracy, this paper proposes the idea of image enhancement based on blind deconvolution combined with noise-adaptive fuzzy equalization algorithms. After image enhancement, the cross-correlation registration algorithm achieves better results. To diagnose a low-temperature corona, the polarization brightness of the red coronal line data and the Mauna Loa Solar Observatory (MLSO)/KCor data are compared to study the structural features and polarization information of the inner corona. We found that the polarization brightness radiation of the red coronal line has overall features similar to that of the MLSO/KCor data, but it has more detailed features, such as the cavity, polar plume, and interplume characteristics. Active regions, polar plumes, and interplumes show similar performance in red coronal line intensity, polarization brightness, degree of linear polarization, K-corona, and electron density along with the radial height. An interesting finding is that the polarization brightness of the red coronal line changes drastically along with the radial height across the cavity above the solar west limb, while neither the K-corona nor the electron density changes obviously. This seems to be a piece of possible evidence indicating that the red coronal line is not unpolarized in the cavity region.