Speckle Noise Reduction Using a New Weighted-Average Filter Based on Euclidean Distance

Abstract

This paper addresses the problem of speckle noise in medical ultrasound imaging. This noise drastically affects the image quality as it is random in nature. For this reason, a new weighted-average filter-based image restoration technique is developed to suppress speckle noise from ultrasound images. The proposed filter structure is based on geometric Euclidean distances of the pixels in an image. The potency of the proposed filter is tested by applying it on a gray-level image and then real ultrasound images using several image quality metrics. A performance comparison with other traditional methods indicates the better performance of the proposed filter.