Thresholding Neural Network (TNN) Based Noise Reduction with a New Improved Thresholding Function
Noorbakhsh Amiri Golilarz, Hasan Demirel · 2017
In this paper, a new thresholding function is introduced for image de-noising in wavelet domain. In this technique we combined the new thresholding function with discrete wavelet transform (DWT). This thresholding function is continuous and nonlinear, so it is applicable on thresholding neural network (TNN) to act as its activation function. Experimental results have shown the superiority of the proposed technique over some alternative methods available in the literature. The proposed method achieves up to 2.90 dB improvement over the state-of-the-art for de-noising ‘Lena’ image.