Hybrid filtering technique to remove noise of high density from digital images

Noise removal is one of the greatest challenges among the researchers, noise removal algorithms vary with the application areas and the type of images and noises. The work proposes a novel hybrid filter which is capable of predicting the best filter for every pixel using neural network and choose the best technique to remove noise with 3x3 mask operation. Proposed algorithm first train the neural network for various filters like mean, median, mode, geometrical mean, arithmetic mean and will use to remove noise later on. Later, the proposed method is compared with the existing techniques using the parameters MAE, PSNR, MSE and IEF. The experimental result shows that proposed method gives better performance in comparison with MF, AMF and other existing noise removal algorithms and improves the values of various parameter