An Atomic Technique For Removal Of Gaussian Noise From A Noisy Gray Scale Image Using LowPass-Convoluted Gaussian Filter

Abstract

During Acquisition of an image using some digital devices we often observe various types of noises. Between them, one of the major types of noise that we often found is the Gaussian noise. When an image contains Gaussian noise, it produces several impurities which are very difficult to detect and eliminate. As a part of the image restoration process, removal of Gaussian noise from a digital image is always a matter of challenge. Before removal of Gaussian noise, we need to convert the digital image into a grayscale image which may contain different percentages of Gaussian noise. Throughout the last few decades, lots of Gaussian noise removal filters or algorithms have been proposed in different international conference papers and acclaimed journals. But a very few of them were successful as far as detecting and eliminating of Gaussian noise from a digital image is concerned. Moreover, these proposed methods also contain several drawbacks and pitfalls which are creating obstacles regarding the generation of the enhanced output image. In this paper, we propose a unique and atomic technique for removal of Gaussian noise from a digital noisy image which is not only capable of detecting and eliminating Gaussian noise, present in the digital image but also capable of generating an enhanced output image. We also try to establish that our proposed method is giving much better result in comparison to other popular filters or algorithms. In order to do that we have invoked a comparative study in experimental results and analysis portion of these paper by calculating PSNR, MSE and RMSE.