摘要
为了提高低质量手指静脉图像的识别率,提出了一种基于细线距离顺序统计量的手指静脉快速识别算法.利用顺序统计量计算部分较小细线距离的平均值,提高了存在伪静脉手指图像识别性能.在细线点集距离计算中采用邻域搜索的方法,相比采用逐点匹配的MHD点集计算方法,缩短了大量的匹配时间.实验结果表明,所提算法比广泛使用的基于特征点的MHD算法具有更好的性能.
In order to improve the recognition performance of the low quality finger vein image, a fast recognition algorithm based on the line distance order statistics is proposed. The algorithm makes use of the order statistics to calculate the average of some small fine line distance, which improves the recognition performance of the finger image included fake vein. The neighborhood search algorithm based on the line distance calculation compared with the MHD point set calculation method matching point by point, which shortens the matching time largely. The experimental results show that the proposed algorithm has better performance than the widely used MHD algorithm based on feature points.
出处
《杭州电子科技大学学报（自然科学版）》
2017年第4期20-24,共5页
Journal of Hangzhou Dianzi University：Natural Sciences
基金
国家自然基金资助项目(61401133)
中国博士后科学基金资助项目(2014M562302)
浙江省新苗人才计划资助项目(2016R407069)
关键词
手指静脉
匹配识别
MHD算法
finger vein
matching recognition
MHD algorithm
