A Neutrosophic based Non-Local Means Filter for Despeckling of Medical Ultrasound Images

Abstract

In this paper, a new weight function based on neutrosophic logic is presented for improving the performance of non-local means (NLM) filter to deal with speckle noise in ultrasound (US) images. In neutrosophic domain, each pixel is characterized by three components including truth membership T, indeterminacy membership I and falsity membership F. In our proposed method, according to the nature of noise in US images, modified functions are introduced for obtaining neutrosophic components. Then, we apply these components for measuring the similarity between pixels and designing a proper weight function to improve despeckling performance of NLM filter. The evaluations on synthetic and real US data show superiority of our proposed method compared to other state-of-the-art techniques.