| Presentation | 2015-03-06 A Study on Learning-Based Super-Resolution By Sparse Coding Using Image Search Based on SSIM Kodai AZUMA, Masayuki KUROSAKI, Hiroshi OCHI, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Super-Resolution is one of the methods used to enhance image resolution. Up until now, an Learning-Based Super-Resolution method that utilizes high resolution image and low resolution image pairings in order to compensate high frequency component has been proposed. In this conventional method, similarity patch is retrieved by using PSNR and differential filter. However, PSNR can only evaluate the average error in brightness while differential filter can only evaluate the difference. Thus, it tends to search for images that are not similar to the structure of the given image. Particularly in Learning-Based Super-Resolution using Sparse Coding, considering feature value as the basis, it is important to search for patches that preserve the most suitable image structure in order to restore the image. In this study, we propose an image searching method using SSIM that considers the image's average brightness, contrast, and similarities in image structure. We show the effectiveness of the proposed method over the conventional method from the results after the proposed super resolution method is conducted. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Sparse Coding / Learning-Based Super-Resolution / SSIM |
| Paper # | SIS2014-107 |
| Date of Issue |
| Conference Information | |
| Committee | SIS |
|---|---|
| Conference Date | 2015/2/26(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Smart Info-Media Systems (SIS) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | A Study on Learning-Based Super-Resolution By Sparse Coding Using Image Search Based on SSIM |
| Sub Title (in English) | |
| Keyword(1) | Sparse Coding |
| Keyword(2) | Learning-Based Super-Resolution |
| Keyword(3) | SSIM |
| 1st Author's Name | Kodai AZUMA |
| 1st Author's Affiliation | Dept. of Computer Science and Electronics, Kyushu Institute of Technology() |
| 2nd Author's Name | Masayuki KUROSAKI |
| 2nd Author's Affiliation | Dept. of Computer Science and Electronics, Kyushu Institute of Technology |
| 3rd Author's Name | Hiroshi OCHI |
| 3rd Author's Affiliation | Dept. of Computer Science and Electronics, Kyushu Institute of Technology |
| Date | 2015-03-06 |
| Paper # | SIS2014-107 |
| Volume (vol) | vol.114 |
| Number (no) | 496 |
| Page | pp.pp.- |
| #Pages | 5 |
| Date of Issue | |