Anisotropic Diffusion-Based Unsharp Masking for Sharpness Improvement in Digital Images
Zohair Al‐Ameen, Mayada A. Al-Healy, Rahma A. Hazim · 2019
Various available imaging systems are capturing images with deficient sharpness due to numerous unavoidable shortcomings. Perceiving and extracting information from such images is uneasy. Hence, it is required to process these images properly to produce sharper and clearer details. Many methods exist that can be used to increase the sharpness of digital images. Among such, the unsharp mask has gained high popularity due to its rapidness and simplicity. Still, this filter usually degrades the processed images by an overshoot effect, which appears around the edges as white shades. In this study, an anisotropic diffusion-based unsharp mask filter, so-called ADUSM, is proposed, in that the degraded image is filtered using an amended anisotropic diffusion filter rather than processing it only by a low-pass Gaussian filter. This modification permitted the attenuation of the overshoot artefact which yielded to obtain better quality results. The ADUSM is tested with several types of images and assessed with two adequate quality metrics. Many experiments indicated that the proposed filter can outperform different existing methods and produce satisfactory results with reasonable application time.