2015 年 135 巻 2 号 p. 81-86
The Iterative Back Projection (IBP) or Learning-based Super Resolution (LBSR) method is used as a substitute for bicubic interpolation in the resolution enhancement process. The IBP method has a limitation that it cannot be used to increase the magnification of an image. Furthermore, if the desired high-resolution information does not exist in the learning database, the LBSR method restores the undesired high-resolution information in instances where the low-resolution image is not available. In this paper, we propose a Super Resolution method that combines Bilateral Back Projection using Local Variance (BBPLV) and the detection of similar structure and data fusion. Through our experimental results, we show the effectiveness of our proposed method.
J-STAGEがリニューアルされました!https://www.jstage.jst.go.jp/browse/-char/ja/