Color Image Denoising via Sparse and Redundant Representations over Online Dictionary
| Autor: | Yi-Sheng Huang, 黃弌聖 |
|---|---|
| Rok vydání: | 2011 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 99 An image is easy to get noise by transferring, so image denoising is to address this problem. Image denoising is an ill-posed problem in image processing. In this study, a color image denoising method using sparse and redundant representations over online dictionary is proposed. The proposed image denoising method contains six stages: (1) separating the noisy color image to the RGB components; (2) color decorrelation; (3) Constructing a dictionary which name called “online dictionary” by online dictionary learning algorithm with two images. (4) denoising the three components by online dictionary separately; (5) color correlation; and (6) merging the denoised RGB components to a denoisied color image. Based on experimental results obtained in this study, the subjective or objective measure for the results, the proposed method provides the better image denoising results compared with two comparison methods. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|