Journal of Computer Applications ›› 2021, Vol. 41 ›› Issue (3): 898-903.DOI: 10.11772/j.issn.1001-9081.2020060965

Special Issue: 前沿与综合应用

• Frontier and comprehensive applications • Previous Articles     Next Articles

Vein recognition algorithm based on Siamese nonnegative matrix factorization with transferability

WANG Jinkai, JIA Xu   

  1. School of Electronics and Information Engineering, Liaoning University of Technology, Jinzhou Liaoning 121001, China
  • Received:2020-07-06 Revised:2020-10-21 Online:2021-03-10 Published:2021-01-15
  • Supported by:
    This work is partially supported by the National Natural Science Foundation of China (61502216), the Guidance Plan of Liaoning Natural Science Foundation (2019-ZD-0700).

基于迁移孪生非负矩阵分解的静脉识别算法

王锦凯, 贾旭   

  1. 辽宁工业大学 电子与信息工程学院, 辽宁 锦州 121001
  • 通讯作者: 贾旭
  • 作者简介:王锦凯(1982-),男,辽宁锦州人,实验师,硕士,主要研究方向:计算机视觉、机器学习;贾旭(1983-),男,辽宁开原人,副教授,博士,CCF会员,主要研究方向:模式识别、机器学习。
  • 基金资助:
    国家自然科学基金资助项目(61502216);辽宁省自然科学基金指导计划项目(2019-ZD-0700)。

Abstract: Concerning the problem that the recognition algorithm which was obtained under one vein image dataset is lack of universality to other datasets, a Siamese Nonnegative Matrix Factorization (NMF) model with transferability was proposed. Firstly, the supervised learning for the vein images with same labels in the source dataset was achieved by using two NMF models with the same structures and the parameter sharing. Then, the vein feature differences between two different datasets were reduced through using maximum mean discrepancy constraint, that is to transfer the knowledge in the source dataset to the target dataset. Finally, the matching of vein images was realized based on cosine distance. Experimental results show that, the proposed recognition algorithm can not only achieve the high recognition accuracy on the source dataset, but also respectively reduce the average False Accept Rate (FAR) and average False Reject Rate (FRR) to 0.043 and 0.055 on the target dataset when using only a small number of vein images in the target dataset. In addition, the average recognition time of the proposed algorithm is 0.56 seconds, which can meet the real-time requirement of recognition.

Key words: vein recognition, Siamese model, Nonnegative Matrix Factorization (NMF), transfer learning, gradient descent algorithm

摘要: 针对某一静脉图像数据集下获得的识别算法对于其他数据集缺少普适性的问题,提出了一种具有迁移性的孪生非负矩阵分解(NMF)模型。首先,通过采用两个结构相同且参数共享的NMF模型实现了对源数据集中带有相同标签静脉图像的有监督学习;然后,通过使用最大均值差异约束降低了不同数据集之间静脉特征的差异性,即将源数据集中的知识迁移至目标数据集中;最后,基于余弦距离实现静脉图像的匹配。实验结果表明,所提的识别算法不仅可以在源数据集下上获得较高的正确识别率,而且仅利用目标数据集中的少量静脉图像便可使得在目标数据集上的平均错误接受率(FAR)与平均错误拒绝率(FRR)分别降低至0.043与0.055。此外,所提算法平均0.56 s的识别时间可以满足识别的实时性要求。

关键词: 静脉识别, 孪生模型, 非负矩阵分解, 迁移学习, 梯度下降法

CLC Number: