随着现代社会在信息技术方面的不断进步和发展，人们对身份的认证与识别的需求在不断增大，而要求也在不断提高。传统的身份认证与识别技术，如口令、密码等在安全性、独特性等方面存在诸多弊端，所以基于生物特征的识别技术近年来越来越凸显其优势。而指静脉诸多生物特征的一类，以其独特的优势在市场中占据了一席之地。然而，作为一类相对新兴的生物特征识别技术，基于指静脉和指背纹的身份认证技术也存在不少可以改进和提高的方面。接下来，本文首先探讨和研究了一种基于新型多模态的指静脉和指背纹特征，主要内容包括：（1）设计与搭建图像采集装置，通过在采集阶段融合指静脉和指背纹纹理，减少了存储的空间和计算的时间，并基于这种多模态纹理建立了一个包含126个融合了指静脉和指背纹纹理的数据库；（2）针对融合的指静脉和指背纹纹理的特点，提出了基于Gabor滤波的交叉点加强编码方法（IGDC），在建立的数据库上取得了良好的性能。之后，为了更加充分的利用指静脉的三维结构特征，本文又进一步探讨了基于多视角的指静脉重建算法，主要内容有：（1）增加了手指角度旋转控制器和遮光罩以适应多视角采集的要求，具有成本相对低廉、采集快捷方便的特点，并建立了一个包含200个手指样本的指静脉多视角图像库。（3）设计与实现了基于多视角图像的投影重建算法3D-SFS，通过重建指静脉的三维拓扑结构，在新建的多视角图像库上相较于传统的二维特征提取算法取得了最优的性能。实验结果表明，基于多模态的指静脉和指背纹的特征相较于传统的单模态特征具有更好的性能，而基于多视角的三维的指静脉特征相较于传统的二维特征，也有相当的优势。

With the continuous progress and development of information technology in modern society, the demand for identity authentication and identification is increasing, so as the requirement. Traditional methods such as password have many disadvantages in security, uniqueness and so on, so the biometric based technology has shown more and more advantages in recent years, such as finger vein. However, as a relatively new biometric identification technology, there are a lot of ways to improve its performance.In the following, this paper firstly discusses a new type of multi-modal finger vein and finger dorsal pattern:(1) We design and construct an image acquisition device, which reduces storage space and computation time by fusing finger vein and finger dorsal texture in image acquisition step, and establish a database containing 126 multi-modal finger samples. (2) We propose the Intersection Enhanced Gabor Based Direction Coding (IGDC) method specific to the fused finger vein and finger dorsal texture, which performs well on the established database.Then, to fully utilize the 3-dimension feature of finger vein, we further discuss multi-view based finger vein reconstruction algorithm, main content includes:(1) We add a finger rotation controller and a shading hood to adapt to meet the requirement of multi-view based acquisition, which is comparatively cheap and convenient to use, and we use the modified device to establish a database containing 200 finger samples. (2) We design and implement the multi-view based 3-dimension shape from silhouette (3D-SFS) algorithm, which constructs the 3-dimension topology structure of finger vein and performs the best on the new database compared with traditional 2D feature extraction methods.As the experiment results show, the multi-modal feature of finger vein and finger dorsal textures performs better compared with traditional unimodal feature, and the multi-view 3-dimension feature of finger vein and finger dorsal textures has advantage compared with the traditional 2-dimension feature.