WEKO3
アイテム
Eye Region Detection by Likelihood Combination for Improving Iris Authentication
http://hdl.handle.net/10458/6455
http://hdl.handle.net/10458/6455136f8d39-b06f-473d-975b-7880eab28b55
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
本文 (923.8 kB)
|
|
| Item type | 紀要論文 / Departmental Bulletin Paper(1) | |||||
|---|---|---|---|---|---|---|
| 公開日 | 2020-06-21 | |||||
| タイトル | ||||||
| タイトル | Eye Region Detection by Likelihood Combination for Improving Iris Authentication | |||||
| 言語 | en | |||||
| 言語 | ||||||
| 言語 | jpn | |||||
| キーワード | ||||||
| 言語 | en | |||||
| 主題Scheme | Other | |||||
| 主題 | Eye region detection, AKAZE, Template matching, OSIRIS | |||||
| 資源タイプ | ||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_6501 | |||||
| 資源タイプ | departmental bulletin paper | |||||
| 著者 |
Thae, Su Tun
× Thae, Su Tun× 椋木, 雅之 |
|||||
| 抄録 | ||||||
| 内容記述タイプ | Abstract | |||||
| 内容記述 | In this research, we propose a likelihood combination method to detect the eye region to improve iris authentication. For face images with noises, iris has more possibility to be wrongly segmented in the segmentation stage. In order to avoid wrong segmentation on other organs from the face, the accurate eye region is detected firstly by our proposed method before iris segmentation. The input face image is divided into left and right half face images in order to detect the individual eyes. The likelihood images are created by AKAZE feature matching and template matching for each half input image. The feature points obtained from AKAZE feature matching are used to create first likelihood image and the matched points from template matching are used to create second likelihood image. Then, these two likelihood images are combined and the eye region is detected based on the highest peak from combined likelihood image. Giving a cut eye region image as input for the iris segmentation stage has more possibility to be correctly segmented by OSIRIS, that is a reference iris authentication system, as it contains iris area. Experimental results show that our likelihood combination method has good performance to detect eye region accurately and detected eye region images can authenticate better than giving face images as input. | |||||
| 言語 | en | |||||
| 書誌情報 |
ja : 宮崎大学工学部紀要 en : Memoirs of Faculty of Engineering, University of Miyazaki 巻 47, p. 331-338, 発行日 2018-07 |
|||||
| 出版者 | ||||||
| 出版者 | 宮崎大学工学部 | |||||
| 言語 | ja | |||||
| 出版者 | ||||||
| 出版者 | Faculty of Engineering, University of Miyazaki | |||||
| 言語 | en | |||||
| ISSN | ||||||
| 収録物識別子タイプ | ISSN | |||||
| 収録物識別子 | 05404924 | |||||
| 書誌レコードID | ||||||
| 収録物識別子タイプ | NCID | |||||
| 収録物識別子 | AA00732558 | |||||
| 著者版フラグ | ||||||
| 出版タイプ | VoR | |||||
| 出版タイプResource | http://purl.org/coar/version/c_970fb48d4fbd8a85 | |||||
