Image restoration by the adaptive fuzzy backpropagation hybrid network

Abstract

In this paper, we introduce a new nonlinear fuzzy neural filter, namely the adaptive fuzzy backpropagation hybrid network (AFBHN), which is shown powerful for image restoration and capable of generalization. The AFBHN restores noisy images even when the impulsive noise rate is over 30%. Under Gaussian white noise testing, the AFBHN has shown better results than the Wiener filter (WF). The generalization capability is shown by performing simulations in which many similar images are given, only one is randomly selected and used for training the AFBHN. The trained network is shown capable of restoring all of the similar images when impulsive noise or Gaussian white noise is present.