Unsharp Masking of Medical Images using Rolling Guidance Filter

Abstract

Medical Images consists of both high frequency and low frequency information needed by Physician for diagnostics and treatment purpose. Since, human eye is sensitive to high frequency information, the enhancement of high frequency information of medical image is very important. Most image enhancement techniques use the traditional unsharp mask filter to enhance the detail and edge information of medical image. While enhancing the detail and edge information, Gaussian Filter in traditional Unsharp Masking Filter introduces halo effects around edges. To remove this halo effects in enhanced images, this paper proposes a novel image enhancement filter in which Gaussian Filter is replaced by rolling guidance filter. This method provides satisfying results over Traditional Unsharp Mask Filter, Amended Unsharp Mask Filter, Anisotropic Diffusion based Unsharp Mask Filter and Guided Filtering based Unsharp Mask Filter in terms of high frequency information measures like mean gradient, spatial frequency and sharpness.