[1]史骏鹏,吴一全.基于混沌蜂群优化的指纹匹配算法[J].智能系统学报,2016,11(5):613-618.[doi:10.11992/tis.201601038]
SHI Junpeng,WU Yiquan.A fingerprint minutiae matching algorithm based on chaotic bee colony optimization[J].CAAI Transactions on Intelligent Systems,2016,11(5):613-618.[doi:10.11992/tis.201601038]
点击复制
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
11
期数:
2016年第5期
页码:
613-618
栏目:
学术论文—机器学习
出版日期:
2016-11-01
- Title:
-
A fingerprint minutiae matching algorithm based on chaotic bee colony optimization
- 作者:
-
史骏鹏1, 吴一全1,2
-
1. 南京航空航天大学 电子信息工程学院, 江苏 南京 211106;
2. 南京理工大学 江苏省社会安全图像与视频理解重点实验室, 江苏 南京 210094
- Author(s):
-
SHI Junpeng1, WU Yiquan1,2
-
1. College of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China;
2. Jiangsu Key Laboratory of Image and Video Understanding for Social Safety, Nanjing University of Science and Technology, Nanjing 210094, China
-
- 关键词:
-
指纹识别; 特征点匹配; 群智能优化; 人工蜂群; 混沌策略; 可变界限盒; 适应度函数; 极坐标
- Keywords:
-
fingerprint recognition; minutiae matching; swarm intelligence optimization; artificial bee colony; chaos strategy; variable boundary box; fitness function; polar coordinates
- 分类号:
-
TP391.4
- DOI:
-
10.11992/tis.201601038
- 摘要:
-
为了进一步加快指纹匹配算法的运算速度、提高识别效率,提出了一种基于混沌蜂群优化和可变界限盒的指纹匹配算法。首先,结合人工蜂群优化算法收敛速度快、控制参数少、能够避免局部最优等优点以及混沌策略的类随机性、高遍历性等特点,在指纹点匹配中引入混沌蜂群优化算法,并设计兼顾了匹配精度和运算时间的适应度函数;然后利用适应度函数估计出指纹特征匹配的几何变换参数并进行指纹点特征的粗匹配;最后,利用可变界限盒进行精匹配,避免指纹图像局部形变带来的影响。大量实验结果表明,与基于局部特征的指纹匹配算法、基于遗传算法优化的指纹匹配算法相比,本文提出的算法所需运算时间更短,匹配精度更高。
- Abstract:
-
In order to further improve the operational speed and the recognition efficiency of fingerprint matching algorithms, a fingerprint matching algorithm based on chaotic bee colony activity and a variable boundary box was proposed. Firstly, by combining the advantages of artificial bee colony optimization including fast convergence times, fewer control parameters, and the lack of local optima, with the features of a chaos strategy including its random-like property and ergodicity, the chaotic bee colony activity was introduced into point pattern matching for fingerprint images. A corresponding fitness function incorporating both matching accuracy and operational time was then designed. The corresponding fitness function was then used to estimate the geometric transformation parameters for fingerprint rough matching. Finally, a variable boundary box can be used for fine matching, because it avoids any influences relating to local deformation of the fingerprint images. A large number of experimental results show that, compared with two alternative fingerprint matching algorithms (based on local features and genetic algorithm optimization, respectively) the proposed algorithm has a shorter operational time and has higher matching accuracy.
备注/Memo
收稿日期:2016-01-28。
基金项目:国家自然科学基金项目(61573183);江苏省社会安全图像与视频理解重点实验室(南京理工大学)开放基金项目(JSKL201302);江苏省高校优势学科建设工程项目(2012).
作者简介:史骏鹏,男,1990年生,硕士研究生,主要研究方向为图像处理与视频通信。发表学术论文3篇;吴一全,男,1963年生,博士,教授,博士生导师,主要研究方向为图像处理与分析、目标检测与识别、智能信息处理。发表学术论文250余篇。
通讯作者:吴一全.E-mail:nuaaimage@163.com
更新日期/Last Update:
1900-01-01