Adaptive Color Image Denoising using Neural Networks

M.H. Sedaaghi and E. Ajami (Iran)


Neural networks, adaptive filtering, color image denoising.


In this paper the authors present a novel approach in color image denoising based on artificial neural networks (ANNs). The main objective of this paper is to design an adaptive noise canceller using appropriate neural networks (NNs). Two groups of NNs, which are recurrent (RNN) and feed-forward (FNN), have been designed and applied as adaptive filters for color image noise removal tasks. Experimental results, illustrate the superb performance of NN-based adaptive filtering. The dominance of the pro posed method over conventional adaptive algorithms, such as LMS (least mean square) , RLS (recursive least square) and KLM (Kalman), is proved for both single and multiple denoising.

Important Links:

Go Back