A Novel Approach for Image Denoising Based on Artificial Neural Networks

This study presents a novel approach based on artificial neural networks (anns) to remove noises from defected images. anns were trained with two different learning algorithms, levenberg-marquardt and extended-delta-bar-delta, for speeding up the training and feedforward calculation processes. the restored results were also compared to the classical techniques, fft, wiener+median filtering and wavelet denoising. the results were shown that the proposed novel neural model provides simplicity and accuracy to remove noises from defected images without estimating any mathematical model than the others.