ABSTRACT
It is well known that many imaging problems require the inversion of a Fredholm integral equation of the first kind. This chapter provides numerical examples which demonstrate that a great deal of improvement in reconstruction quality can be realized by varying the amount of regularization throughout the interval of interest. There are many directions this idea may be taken to generalize. For example, the regularization operator rather than just the regularization parameter could be varied on each component of the partition. The partition itself can be varied, and varying the components might lead to significant improvement. The partition can be refined. More than two components could be tried, although with each new component comes an additional variable.
