Palm vein recognition technology offers advantages in terms of security and privacy. However, traditional methods rely on predefined rules about hand shape and positioning, affecting user experience. This paper focuses on the research of ROI extraction algorithm for palm vein recognition, utilizing an improved SSD [1] network to achieve multiple tasks such as palm classification, ROI localization, and gesture correction. It enhances the localization accuracy under various scenarios, including complex backgrounds and special hand shapes. The proposed approach in this paper enables non-contact palm recognition technology, implementing 360-degree omnidirectional recognition, thereby improving convenience and feasibility in its usage.

Authors and Affiliations

1. Shanghai Jiao Tong University, No. 800 Dongchuan Road, Minhang District, Shanghai, 200240, China

   Xu Chen & Yang Genke

Corresponding author

Correspondence toYang Genke.

Editors and Affiliations

1. Hefei University of Technology, Hefei, China

   Wei Jia

2. South China University of Technology, Guangzhou, China

   Wenxiong Kang

3. China University of Mining and Technology, Xuzhou, China

   Zaiyu Pan

4. Shandong University, Jinan, China

   Xianye Ben

5. China University of Mining and Technology, Xuzhou, China

   Zhengfu Bian

6. Southern University of Science and Technology, Shenzhen, China

   Shiqi Yu

7. Chinese Academy of Sciences, Beijing, China

   Zhaofeng He

8. China University of Mining and Technology, Xuzhou, China

   Jun Wang

Reprints and permissions

© 2023 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Policies and ethics