Improved Orthogonal Fractal Super-Resolution Using Range Adjustment and Domain Extension
スポンサーリンク
概要
- 論文の詳細を見る
This paper presents a modified orthogonal fractal super-resolution (OFSR) method to improve the visual quality of an image along sharp edges. Although the OFSR method constructs a high-quality high-resolution image from a low-resolution counterpart, there are ringing artifacts observed along sharp edges which make the visual quality relatively low with respect to the numerical quantity. These artifacts are mainly caused by unnecessarily exaggerated pixel contrast along sharp edges within a range block. We restrict each contracted pixel value in a range block to a value between the minimum and maximum of its domain block pixel values. We also extend the domain block of the contraction function and find a better domain block using the range block mean. At the final step of the iteration, we adjust each pixel in the range block so that the range block mean and the corresponding pixel value of the low-resolution image are equal. According to our experimental results, the proposed method improves the visual quality along sharp edges and shows higher levels of numerical quantity than the OFSR method.
著者
-
WEE Youngcheul
Department of Information and Computer Engineering, Ajou University
-
CHOI Jung-Ju
Department of Digital Media, Ajou University
-
LEE Moojae
Department of Information and Computer Engineering, Ajou University
関連論文
- On Improving JPEG Entropy Coding by means of Sub-Stream Extraction
- Improved Orthogonal Fractal Super-Resolution Using Range Adjustment and Domain Extension