TS-Net: Texture and Shape Stream Network for Retinal Vessel Segmentation
文献类型：会议论文
| 作者 | Wang, Ying2; Fan HJ(范慧杰)1,3 ; Yang, Dawei2
|
| 出版日期 | 2021 |
| 会议日期 | November 26-28, 2021 |
| 会议地点 | Shanghai, China |
| 页码 | 667-671 |
| 英文摘要 | Retinal blood vessels are the only blood vessels that can be directly observed in the human body through non-invasive methods. The precise segmentation of retinal blood vessels helps us diagnose fundus diseases. We propose a network model-TS-Net (Texture and Shape Stream Net) that combines shape information and texture information. The shape stream is used to capture the shape and orientation of the vascular tree. Shape stream is composed of Residual Block, Gated Convolutional Layers (GCL) and vascular orientation modules. The texture stream is used to learn dense pixel information and features. In order to prevent the network from overfitting, we modify the convolutional layer network and get the Improved Dropout Block. We fuse the dense pixel information features from the texture stream and the shape trend features from the shape stream to produce a refined segmentation result. We use the color fundus DRIVE dataset and the CHASE_DB1 dataset to evaluate TS-Net. Through the comparison of experimental results, we find that TS-Net is effective and achieves advanced performance. |
| 产权排序 | 2 |
| 会议录 | 2021 27th International Conference on Mechatronics and Machine Vision in Practice, M2VIP 2021
 |
| 会议录出版者 | IEEE |
| 会议录出版地 | New York |
| 语种 | 英语 |
| ISBN号 | 978-1-6654-3153-8 |
| 源URL | [http://ir.sia.cn/handle/173321/30500]  |
| 专题 | 沈阳自动化研究所_机器人学研究室 |
| 通讯作者 | Wang, Ying |
| 作者单位 | 1.Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences, Shenyang 110169, China 2.Shenyang Ligong University, School of Information Technology and Engineering, Liaoning, Shenyang 110159, China 3.State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences China, Shenyang 110016, China |
| 推荐引用方式 GB/T 7714 | Wang, Ying,Fan HJ,Yang, Dawei. TS-Net: Texture and Shape Stream Network for Retinal Vessel Segmentation[C]. 见:. Shanghai, China. November 26-28, 2021. |
入库方式： OAI收割
来源：沈阳自动化研究所
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。