This paper describes a set of algorithms that enable virtually complete ring artifact removal from tomographic imagery
with minimal to negligible contamination of the underlying data. These procedures were created specifically to deal
with data as acquired at the University of Texas high-resolution X-ray CT facility, but are likely to be applicable in other
settings as well. In most cases corrections are optimally applied to sinogram data before reconstruction, but a variant is
developed for correcting already-reconstructed images. The algorithms make particular use of repetitive aspects of the
artifact across images to improve behavior. However, fully utilizing this functionality requires processing entire data
sets simultaneously, rather than one image at a time. A number of parameters may be adjusted to optimize results for
particular data sets.