首页 | 本学科首页   官方微博 | 高级检索  
     

人体手背静脉血管图像增强与特征匹配
引用本文:吕佩卓,赖声礼. 人体手背静脉血管图像增强与特征匹配[J]. 计算机工程与设计, 2007, 28(15): 3623-3625,3635
作者姓名:吕佩卓  赖声礼
作者单位:华南理工大学,电子与信息学院,广东,广州,510640;华南理工大学,电子与信息学院,广东,广州,510640
摘    要:为提高手背静脉血管图像识别的准确性和改善图像采集的灵活性,提出了一种新的手背静脉图像识别算法.使用自适应直方图均衡法增强背景和静脉纹理图像的总体对比度和局部细节,使用分水岭算法得到静脉纹理图像的单像素级图像,最后采用纹理相似性原理进行特征匹配和识别.试验结果证明,该算法在图像采集位置灵活变动情况下的错误接受率为0%,错误拒绝率4.6%,利用此算法进行手背静脉血管图像识别是可行的,具有良好的应用前景.

关 键 词:静脉识别  自适应直方均衡化  分水岭算法  特征提取与匹配  相似性原理
文章编号:1000-7024(2007)15-3623-03
修稿时间:2006-09-12

Image enhancement and feature matching of palm-dorsa vein
L Pei-zhuo,LAI Sheng-li. Image enhancement and feature matching of palm-dorsa vein[J]. Computer Engineering and Design, 2007, 28(15): 3623-3625,3635
Authors:L Pei-zhuo  LAI Sheng-li
Affiliation:School of Electronic and Information Engineering, South China University of Technology, Guangzhou 510640, China
Abstract:To improve the veracity of palm-dorsa vein pattern identification and the maneuverability of image collecting,a new algorithm about palm-dorsa vein image identification is proposed.Adaptive histogram balancing is used to enhance the contrast of the background and the vein texture image,and watershed transform is adopted to produce the single pixel image of vein texture and texture comparability theory is used to carry out image matching and identification.Experimental results show that the FAR and FRR are 0% and 12% respec-tively,and it is feasible to use the algorithm to carry out palm-dorsa vein pattern identification.
Keywords:vein identification  adaptive histogram balancing  watershed transform  feature extracting and matching  comparability theory
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号