| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 411458 | Neurocomputing | 2016 | 9 Pages |
Many practical images captured from X-ray systems not only are corrupted by noise, but also are of non-uniformity and low-luminance. Enhancing these images by noise removal, gray-level transformation and/or contrast enhancement, is thus necessary and crucial before further processing.In this paper, a new scheme of enhancement method for X-ray image is presented, which consists of a fuzzy noise removal method and a homomorphic filtering method. The fuzzy noise removal method is composed of fuzzy noise detection and noise filtering. Region detection and degree detection are presented for fuzzy noise detection, and a modified weight median filter is proposed for filtering noise. Then the high-pass enhancement filter is introduced for image sharpening in the proposed homomorphic filtering method. A variety of degraded images and a real X-ray images have been handled by the proposed method, as well as the several existing methods. The comparisons of those experimental results show that the performance of our method overwhelms those of the existing ones.
