Multitask deep active contour-based iris segmentation for off-angle iris images
文献类型：期刊论文
| 作者 | Lu, Tianhao3; Wang, Caiyong1 ; Wang, Yunlong2; Sun, Zhenan2
|
| 刊名 | JOURNAL OF ELECTRONIC IMAGING
 |
| 出版日期 | 2022-07-01 |
| 卷号 | 31期号:4页码:21 |
| 关键词 | iris recognition iris segmentation off-angle iris image active contour attention mechanism |
| ISSN号 | 1017-9909 |
| DOI | 10.1117/1.JEI.31.4.041211 |
| 通讯作者 | Wang, Caiyong(wangcaiyong@bucea.edu.cn) ; Sun, Zhenan(znsun@nlpr.ia.ac.cn) |
| 英文摘要 | Iris recognition has been considered as a secure and reliable biometric technology. However, iris images are prone to off-angle or are partially occluded when captured with fewer user cooperations. As a consequence, iris recognition especially iris segmentation suffers a serious performance drop. To solve this problem, we propose a multitask deep active contour model for off-angle iris image segmentation. Specifically, the proposed approach combines the coarse and fine localization results. The coarse localization detects the approximate position of the iris area and further initializes the iris contours through a series of robust preprocessing operations. Then, iris contours are represented by 40 ordered isometric sampling polar points and thus their corresponding offset vectors are regressed via a convolutional neural network for multiple times to obtain the precise inner and outer boundaries of the iris. Next, the predicted iris boundary results are regarded as a constraint to limit the segmentation range of noise-free iris mask. Besides, an efficient channel attention module is introduced in the mask prediction to make the network focus on the valid iris region. A differentiable, fast, and efficient SoftPool operation is also used in place of traditional pooling to keep more details for more accurate pixel classification. Finally, the proposed iris segmentation approach is combined with off-the-shelf iris feature extraction models including traditional OM and deep learning-based FeatNet for iris recognition. The experimental results on two NIR datasets CASIA-Iris-off-angle, CASIA-Iris-Africa, and a VIS dataset SBVPI show that the proposed approach achieves a significant performance improvement in the segmentation and recognition for both regular and off-angle iris images. |
| WOS关键词 | RECOGNITION ; NETWORK ; NET |
| 资助项目 | National Natural Science Foundation of China[62106015] ; National Natural Science Foundation of China[U1836217] ; National Natural Science Foundation of China[62006225] ; National Natural Science Foundation of China[62071468] ; National Natural Science Foundation of China[61906199] ; Beijing University of Civil Engineering and Architecture Research Capacity Promotion Program for Young Scholars[X21079] |
| WOS研究方向 | Engineering ; Optics ; Imaging Science & Photographic Technology |
| 语种 | 英语 |
| WOS记录号 | WOS:000848751400011 |
| 出版者 | SPIE-SOC PHOTO-OPTICAL INSTRUMENTATION ENGINEERS |
| 资助机构 | National Natural Science Foundation of China ; Beijing University of Civil Engineering and Architecture Research Capacity Promotion Program for Young Scholars |
| 源URL | [http://ir.ia.ac.cn/handle/173211/50020]  |
| 专题 | 自动化研究所_智能感知与计算研究中心 |
| 通讯作者 | Wang, Caiyong; Sun, Zhenan |
| 作者单位 | 1.Beijing Univ Civil Engn & Architecture, Sch Elect & Informat Engn, Beijing, Peoples R China 2.Chinese Acad Sci, Inst Automat, Ctr Res Intelligent Percept & Comp, Natl Lab Pattern Recognit, Beijing, Peoples R China 3.Hunan Univ Technol, Zhuzhou, Hunan, Peoples R China |
| 推荐引用方式 GB/T 7714 | Lu, Tianhao,Wang, Caiyong,Wang, Yunlong,et al. Multitask deep active contour-based iris segmentation for off-angle iris images[J]. JOURNAL OF ELECTRONIC IMAGING,2022,31(4):21. |
| APA | Lu, Tianhao,Wang, Caiyong,Wang, Yunlong,&Sun, Zhenan.(2022).Multitask deep active contour-based iris segmentation for off-angle iris images.JOURNAL OF ELECTRONIC IMAGING,31(4),21. |
| MLA | Lu, Tianhao,et al."Multitask deep active contour-based iris segmentation for off-angle iris images".JOURNAL OF ELECTRONIC IMAGING 31.4(2022):21. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。