Comments on 'Restoration of noisy images modeled by Markov random fields with Gibbs distributions' (and reply)

摘要

For the original article see ibid., vol.36, no.6, p.884-90 (1989). The commenter points out a number of mistakes and misstatements in the above-titled paper by M. Shridhar et al. They concern the relationships of one of the methods described to other methods; the performance of some of the methods; incompleteness of a list of cliques; and statements about various potentials. The author agrees with many of the commenter's points and clarifies others.