Fast digital locally monotonic regression

Locally monotonic regression is the optimal counterpart of iterated median filtering. In a previous paper, Restrepo and Bovik (see ibid., vol.41, no.9, p.2796-2810, 1993) developed an elegant mathematical framework in which they studied locally monotonic regressions in R^N. The drawback is that the complexity of their algorithms is exponential in N. We consider digital locally monotonic regressions, in which the output symbols are drawn from a finite alphabet and, by making a connection to Viterbi decoding, provide a fast O(|A|²αN) algorithm that computes any such regression, where |A| is the size of the digital output alphabet, a stands for lomo degree, and N is the sample size. This is linear in N, and it renders the technique applicable in practice