Maximum bounded entropy: application to tomographic reconstruction

An investigation is conducted into the usefulness of the Maximum Bounded Entropy restoring algorithm which incorporates both lower and upper bounds in the unknown object to yield maximum probable estimates under four conditions. These encompass: (1) image photons behaving as classical particles; (2) a presumed bias of the object toward a flat, grey scene in the absence of image data; (3) object modeling in terms of high gradient foreground details above a smoothly varying backdrop that is to be separately estimated; and (4) Poisson image noise. The resulting estimator obeys the sum of maximum entropy for the occupied photon sites in the object, and maximum entropy for the unoccupied sites. The algorithm is presently applied to the reconstruction of rod cross sections from tomographic viewing.