Image Denoising Algorithms: A Comparative Study of Different Filtration Approaches Used in Image Restoration

To send visual digital images is a major issue in the modern data communication network. The image sent from sender end may not be the same at the receiving end. The image obtained after transmission is often corrupted with noise. The image received at the receiving end needs processing before it can be used for further applications. Image denoising is a process which involves the manipulation of the image data to produce a visually high quality image. In the present work the authors tried to explore the different image denoising techniques and the merits and demerits of all those methods. The authors have discussed different denoising techniques such as filtering approach, wavelet based approach, and multifractal approach and a comparative study of all these methods. The authors have discussed the different additive noise models and also multiplicative models such as Gaussian noise, salt and pepper noise, speckle noise and Brownian noise. Depending on the noise present in an image a particular algorithm is to be selected. When the image is corrupted with salt and pepper noise then it is found that the filtering approach is the best. In case of Gaussian noise the wavelet based approach is found the best denoising method. For any complex type of noise it is found that the multifractal approach is the best method.