Fuzzy logic based computational model for speckle noise removal in ultrasound images

High level of uncertainty is always present due to impulsive noise in ultrasonic images, which may put negative effect on image interpretation, quantitative measurement and diagnostic purposes. In order to deal with this uncertainty, fuzzy modelling is being used which is very helpful in distinguishing noise from edges and other critical details present in the image. In this study, a novel fuzzy logic based non-local mean filter is proposed to model the speckle noise and to restore the degraded image using Fuzzy Uncertainty Modelling (FUM), smoothed by local statistic based information while preserving the image details for low and highly speckled ultrasound images. Proposed denoising technique acquires the local parameters to find distinct similar and non-similar non-local regions using FUM. These homogenous regions are first smoothed through local statistical information and then used to restore the selected noisy pixels using fuzzy logic based noise removal process. The study evaluates the performance of the proposed technique on different real and simulated data sets, and compares the numerical values with existing state of art filters using standard well known global quantitative measure like signal to noise ratio (SNR) and a local error measure - structural similarity index measure (SSIM). Visual and quantitative results demonstrate that the proposed technique outperforms the existing state of the art filters in removing speckle noise while preserving the edges and other important details present in the image.