Noise Smoothing- Improving Image Filtering Methodology

Abstract

Digital images are prone to a variety of types of noise. Noise is the result of errors in the image acquisition process that result in pixel values that do not reflect the true intensities of the real scene. There are several ways that noise can be introduced into an image, depending on how the image is created. For example if the image is under scanner from photography made on a film, noise can be he result of damage occured to the film during data acquisition.Noise occurance can also be there as image is acquired directly in digital form or noise data can also arise due to electronic transmission.Our goal in image restoration is to remove the noise from the image in such a way that original image should not be disturbed. We can use linear filtering method  remove certain   type of images, certain filters such as averaging or Gaussian filters were appropriate for noise filteing purpose.To remove salt and pepper noise from an image using an average filter , Double Derivative filter or a median filter are allowed for comparison.Here a new fuzzy logic method is applied for better efficient performance with less computational techniques as well as with less degradation.