Trimmed median filter for removal of noise from medical image

Abstract

Medical images are very sensitive in nature. During transmission, noise may be added to the image which affects the performance of the system. Generally, images are affected by speckle, Gaussian, and impulse noise. It is essential to remove the noise present in the image. Filters such as adaptive median filter, decision-based algorithm, modified quaternion vector filter, progressive switching median filter, and decision-based unsymmetrical trimmed variant are used for reduction of impulse noise present in the image. A method has been proposed in this paper for removing impulse noise present in the medical images. This technique uses a nonoverlapping window of size 2 × 2, and processing of image is done at the window level by taking consideration of noise-free pixel present in this window. The proposed filter removes impulse noise effectively from medical images.