Derivation of fixed-interval smoothing algorithm using covariance information in distributed parameter systems (Q2495998)

The authors propose a novel recursive least mean squared error fixed interval smoothing algorithm in distributed parameters systems when observations are affected by Gaussian white noise, and the state space model of the signal is not known. A number of theorems and lemmas is proposed. It is also mentioned that the algorithm only requires the information provided by the second order moments of the signal and additive observation noise. Practical experiments are performed on image degraded system for validation of the established theory.