Study of Image Denoising using Wavelet Transform

Visual information transmitted in the form of images has become a main source of communication in present days, usually image taken after transmission is mostly contain noise with them. The image received needs to be processed before it should be used for different useful applications. Image denoising is the process of handling the image data to give a high quality image in terms of visualization. Selection of the denoising algorithm is application dependent. Hence, it is necessary to have knowledge about the noise present in the image so as to select the appropriate denoising algorithm. This paper study the existing denoising algorithms, such as wavelet based approach, filtering approaches, and perform their comparative study and improvement in existing wavelet algorithm with the help of modified techniques. For the different noise models which includes additive and multiplicative types which comprises salt and pepper noise, Gaussian noise, speckle noise and Brownian noise. The filtering approach is considered as best for studying the images which is corrupted with salt and pepper noise. The wavelet based approach proves to be best for denoising images corrupted with Gaussian noise. Cases where characteristics of noise remains complex, the modified approach can be used. A quantitative measure of comparison is provided for different parameter.