摘要
搭建了基于移动摄像头的在线掌纹识别系统,该系统利用提出的自适应掌纹感兴趣区域(Region Of Interest,ROI)提取算法提取掌纹ROI图像,同时采用有限Radon变换(Modified Finite Radon Transform,MFRAT)算法提取掌纹ROI图像的主纹线特征图像,并采用一种基于方向最近邻域(Direction and Nearest Neighbor,DNN)搜索的迭代最近(Iterative Closest Points,ICP)算法对掌纹特征图像进行配准,该配准方法分为粗配准与精配准两个过程。实验表明,该系统能有效克服背景、光照和掌纹ROI图像的旋转平移对掌纹识别精度的影响,同时也具有较高的识别效率。
A method for palmprint recognition on mobile device is proposed. An adaptive algorithm is proposed for ROI extraction of palmprint which is robust to complicated background and unstable light condition. Modified Finite Radon Transform （MFRAT） is used to extract the principal lines. A coarse- to-fine strategy is suggested to register principle lines based on iterative closest points （ICP） and directional nearest neighbor（ DNN ） methods, respectively. Experimental results demonstrate that this method achieves better performance in comparison with other state-of-art palmprint recognition algorithms with respect to accuracy and robustness.
出处
《北京信息科技大学学报（自然科学版）》
2015年第1期70-74,84,共6页
Journal of Beijing Information Science and Technology University
基金
北京市科学技术研究院"创新团队计划"
关键词
自适应掌纹ROI提取
图像配准
搜索的迭代最近算法
竞争编码
adaptive extraction of palmprint＇ s ROI
image registration
iterative closest points
competitive coding
