互连网+应用模式下，对身份识别的需求日益紧迫。由于图像采集设备的多样性，地域环境变化，用户姿态等因素，基于手背静脉的身份识别很难达到高准确率，如何实现鲁棒手背静脉身份识别成为目前理论和应用的关键。.本项目研究弱约束条件下的手背静脉识别技术，主要包括跨设备、跨地域、跨姿态等情况下的识别，拟在高可区分性、高鲁棒和高效率特征提取，高精度匹配分类方法，大样本“深度学习”和小样本“概念学习”的手背静脉识别方法以及基于生物医学信号的数据防伪方向展开深入研究，突破弱约束条件下即跨设备、跨地域、跨姿态条件下手背静脉鲁棒身份识别关键技术，提出弱约束条件手背静脉识别的理论和方法，解决手背静脉识别应用的瓶颈问题，提供手背静脉识别从受控条件走向实际应用环境的跨越式发展的理论基础，相关科学研究工作对深入推动手背静脉识别技术理论研究和应用研究具有重要意义，对于其他弱约束条件下的生物特征识别理论研究具有参考意义。

With the application mode of internet, biometrics is more important. Hand-dorsa vein recognition cannot reach higher accuracy because of different acquisition devices, regional and environmental change, user’s gesture. How to improve the accuracy of hand-dorsa vein recognition, the robustness of recognition is the key problem in spite of theory and application.. This project studies the method of hand-dorsa vein recognition under weak constraint. The weak constraint means across devices, across regions and across gesture. The main research fields include “Distinctive Efficient Robust Features’ Extraction”, “High Accuracy Matching”, “Deep Learning” with a lot of samples, and “Concept Learning” with limited samples method for hand-dorsa vein recognition, “Anti-Fake” based biomedical signal. Robust hand-dorsa vein recognition technology should be studied. Under weak constraint, theory and method about hand-dorsa vein recognition would be put forwarded. In real application, hand-dorsa vein recognition method would be adopted without constraint. The research work of this project will play a more important role in the field of biometrics.