Noise estimation in medical images based on fast discrete curvelet transform via wrapping

Abstract

These days, image processing is a developing field of study. Images have many contribution to research in a number of areas, including biomedical, security, education, and space. Digital images are inherently noisy during the processes of acquiring, coding, transmitting, and processing and the significant barrier is the problem of image corrupted brought on by noise. The two primary causes of the noise are either the transmission procedure from one location to another or the acquisition process itself. Denoising and categorising noise are crucial components of image analysis in medical field. However, there are numerous methods for modifying the image data in order to eliminate noise and restore image quality. A quick overview of the main types of noise is presented in this paper. The process of estimating noise and filtering to produce improved medical images is covered by the proposed framework. In this piece of work, several and various kinds of noise are estimated and detected: Gaussian noise, white noise, Brownian noise, salt-and-pepper, periodic and speckle noises.The proposed system reduces the noising factor in medical images based upon fast discrete curvelet transform (FDCT) via wrapping. MSE has been calculated between original and recovered image.