Image recovery using a new nonlinear adaptive filter based on neural networks

This work defines a new nonlinear adaptive filter based on a feed-forward neural network with the capacity of significantly reducing the additive noise of an image. Even though measurements have been carried out using X-ray images with additive white Gaussian noise, it is possible to extend the results to other type of images. Comparisons have been carried out with the Weiner filter because it is the most effective option for reducing Gaussian noise. In most of the cases, image reconstruction using the proposed method has produced satisfactory results. Finally, some conclusions and future work lines are presented