Restoration of Images Corrupted by Mixed Gaussian Impulse Noise with Weighted Encoding


different types of communication. Data can be text, image, audio and video. During the image transmission the images are affected by many types of noise. Mainly Additive White Gaussian Noise (AWGN), Impulse Noise (IN) and combination of both called as “mixed noise”. Removal of mixed noise from the original image is critical and challenging work. The noise spreading is not having any predefined model consisting of heavy tail, due to which quality of image reduces. To remove the mixed noise from the image, many methods exist. These are detection based methods. In this method, locations of the noise are detected and then from these locations noise is removed using some algorithms, based on intensity and amount of noise. But these methods will give poor result, if the mixed noise is strong. Hence this paper implements a new effective method to remove the mixed noise. In this method there is no separate step for detection of different types of noise, instead pixel detection via weighted encoding is done which deals with AWGN and IN simultaneously. This proposed method performs better than existing image denoising methods. It can be applied with multiple types of IN and even in the condition when noise content is more.

Cite this paper

@inproceedings{Bhat2015RestorationOI, title={Restoration of Images Corrupted by Mixed Gaussian Impulse Noise with Weighted Encoding}, author={Om Prakash V. Bhat and Shrividya G and Nagaraj N. S}, year={2015} }