Multilevel-DWT-Based Image Denoising Using Adaptive Neuro-Fuzzy Inference System

摘要

Images corrupted by noise requires enhancement for subsequent processing. Traditional approaches of denoising rely upon spatial, statistical, and spectral properties of image which at times fails to capture the finite details. Discrete wavelet transform (DWT) is a commonly adopted method for image processing applications. Fuzzy-based systems are suitable for modeling uncertainty. In the proposed work, we present a hybrid approach which combines multilevel DWT and adaptive neuro-fuzzy inference system (ANFIS) to capture the benefits of two different domains into a single framework. We apply our algorithm to denoise the images corrupted by multiplicative noise like speckle noise. The results obtained shows that the proposed method proves effective for denoising of images.