您的位置:山东大学 -> 科技期刊社 -> 《山东大学学报(工学版)》

山东大学学报(工学版) ›› 2012, Vol. 42 ›› Issue (3): 6-12.

• 机器学习与数据挖掘 • 上一篇    下一篇

手指静脉图像感兴趣区域提取方法研究

杨金锋,张海金   

  1. 中国民航大学天津市智能信号与图像处理重点实验室, 天津 300300
  • 收稿日期:2011-11-09 出版日期:2012-06-20 发布日期:2011-11-09
  • 通讯作者: 张海金(1985- ),山东菏泽人,硕士研究生,主要研究方向为手指静脉识别.E-mail:1153766431@qq.com E-mail:1153766431@qq.com
  • 作者简介:杨金锋(1971- ),河南淮阳人,副教授,博士,主要研究方向为生物识别技术.E-mail: jinfengyang@cauc.edu.cn
  • 基金资助:

    国家自然科学基金资助项目(61073143)

Study of region of interest extraction from finger-vein images

YANG Jin-feng, ZHANG Hai-jin   

  1. Tianjin Key Laboratory for Advanced Signal Processing, Civil Aviation University of China, Tianjin 300300, China
  • Received:2011-11-09 Online:2012-06-20 Published:2011-11-09

摘要:

研究了一种手指静脉感兴趣区域(ROI)的定位与提取方法,以解决手指在采集静脉图像时的随机运动对手指静脉识别结果的影响。通过计算手指轮廓线的主方向和末节关节腔位置对手指摆放位置的随机变化进行校正,并初步定位分割出手指静脉感兴趣区域。运用迭代优化方法实现手指静脉感兴趣区域分割结果的精确提取。在小型数据集上进行实验的结果表明:该方法得到的感兴趣区域具有较好的聚类特性。说明该方法能够有效地从同一手指姿态变化的多幅静脉图像中定位并分割出相似度较高的感兴趣区域。

关键词: 手指静脉, 手指轮廓, 关节定位, 感兴趣区域

Abstract:

In order to avoid the impact to the region of interest (ROI) extracted results by random movements of finger when finger-vein images are collected, an approach for locating and extracting the region of interest from finger-vein image was studied. First, the random variations of the finger’s positions were corrected by computing its profile’s primary direction and inter-phalangeal joint’s position, and the ROI was roughly found. Then, an iteration method was applied to obtain the accurate extraction of ROI. The experimental results on a small database showed that the clustering property of ROI extracted better results. This method could extract similar ROIs from fingervein images of a finger, even if its posture is changed.

Key words:  finger-vein image, finger profile, finger joint localization, region of interest

[1] 叶子云,杨金锋. 一种基于加权图模型的手指静脉识别方法[J]. 山东大学学报(工学版), 2018, 48(3): 103-109.
[2] 包书哲1,2,朱月澴1,王春立1*. 基于fMRI的图像底层特征关注研究[J]. 山东大学学报(工学版), 2014, 44(1): 24-28.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!