论文:2013,Vol:31,Issue(5):770-773
引用本文:
朱亚辉, 彭国华, 郑玉山. 核Fisher判别分析在红外与可见光融合评价的应用[J]. 西北工业大学
Zhu Yahui, Peng Guohua, Zheng Yushan. Synthesiszed Performance of Infrared and Visible Image Fusion Based on Kernel Fisher Discriminant Analysis(KFDA)[J]. Northwestern polytechnical university

核Fisher判别分析在红外与可见光融合评价的应用
朱亚辉1, 彭国华1, 郑玉山2
1. 西北工业大学 理学院;
2. 西北工业大学 自动化学院, 陕西 西安 710072
摘要:
针对传统的红外与可见光图像融合效果评价方法存在的问题,提出了核Fisher判别分析(KFDA)的红外与可见光图像融合效果评价方法。将空间信息、图谱信息、噪声抑制、边缘保持度四个方面建立的客观评价指标作为判别因子,采用灰色关联分析划分200组数据为学习样本的训练和检验,利用核Fisher判别技术将低维数据空间的非线性分类问题转化为高维特征空间的线性分类问题,建立快速稳定的多类核Fisher判别分类器,实现了红外与可见光图像融合效果的KFDA评价模型。结果表明:引入核技术的Fisher算法,大大提高了红外与可见光图像融合效果评价的准确性。
关键词:    图像融合    效果评估    核Fisher判别分析    综合方法   
Synthesiszed Performance of Infrared and Visible Image Fusion Based on Kernel Fisher Discriminant Analysis(KFDA)
Zhu Yahui1, Peng Guohua1, Zheng Yushan2
1. Department of Applied Mathematics, Northwestern Polytechnical University, Xi'an 710072, China;
2. Department of Automatic Control, Northwestern Polytechnical University, Xi'an 710072, China
Abstract:
Aiming at the problems of the traditional method of assessing infrared and visible image fusion,we propose synthesized performance based on KFDA. In building this model,discrimination factors are established respectively by considering four aspects of spatial information,spectral information,noise reduction and edge retention,and 11 sets of measured data are divided into training and testing samples with grey correlaiton degree. By using KFDA technology,a nonlinear classification problem of the original data space can be converted into a linear one in feature space of high dimension,which realizes quality estimation of infrared and visible image fusion. The calculated results and their analysis show preliminarily that the KFDA improves the recognition accuracy of the image fusion.
Key words:    calculations    discriminant analysis    image fusion    KFDA(Kernel Fisher Discriminant Analysis)    performance evaluation    synthesis method   
收稿日期: 2013-03-01     修回日期:
DOI:
基金项目: 国家自然科学基金(61070233)资助
通讯作者:     Email:
作者简介: 朱亚辉(1981-),女,西北工业大学博士研究生,主要从事图像融合、图像融合效果评价的研究。
相关功能
PDF(707KB) Free
打印本文
把本文推荐给朋友
作者相关文章
朱亚辉  在本刊中的所有文章
彭国华  在本刊中的所有文章
郑玉山  在本刊中的所有文章

参考文献:
[1] 周浦城, 王 峰, 崔逊学等.基于逼近理想解排序的图像融合效果评价方法.系统工程与电子技术, 2011, 33 (3):681-684 Zhou Pucheng, Wang Feng, Cui Xunxue, et al.Performance Evaluation Approach to Image Fusion Based on Ordering Approximate to Ideal Solution.Syestems Engineering and Electronics, 2011, 33(3): 681-684 (in Chinese)
[2] 何贵青, 陈世浩, 田 云等.多传感器图像融合效果综合研究.计算机学报, 2008, 31 (3): 486-492 He Guiqing, Chen Shihao, Tian Yun, et al.Synthesis Performance Evaluation of Multi-Sensor Image Fusion.Chinese Journal of Computers, 2008, 31 (3): 486-492 (in Chinese)
[3] 徐宝昌, 陈 哲.基于模糊 Choquet 积分的图像融合效果评价.光电工程, 2004, 31 (11): 42-46 Xu Baochang, Chen Zhe.Image Fusion Effect Evaluation Based on Fuzzy Choquet Integration.Opto-Electronic Engineering,2004, 31 (11): 42-46 (in Chinese)
[4] 胡春海, 丁志秀.基于模糊算法的多聚焦图像融合效果评价研究.传感器学报, 2007, 20(4): 885-888 Hu Chunhai, Ding Zhixiu.Multi-Focus Image Fusion Effect Evaluation Based on Fuzzy Algorithm.Chinese Journal of Sensors and Actuators, 2007, 20(4): 885-888 (in Chinese)
[5] 薛新美, 王敬东, 李 鹏等.基于神经网络的图像融合效果综合评价.红外技术, 2006, 28(12): 699-703 Xue Xinmei, Wang Jindong, Li Peng, et al.Comprehensive Evaluation of Image Fusion Based on Artificial Neural Network.Infrared Technology, 2006, 28(12): 699-703 (in Chinese)
[6] Mika S, Ratsch G, Weston J, et al.Fisher Discriminant Analysis with Kernels.Neural Networks for Signal Processing IX,1999.Proceedings of the 1999 IEEE Signal Processing Society Workshop, 1999, 41-48
[7] 朱卫纲, 周荫清, 徐华平等.遥感图像融合效果客观评价.火力与指挥控制, 2010, 35(2): 2-5 Zhu Weigang, Zhou Yinqing, Xu Huaping, et al.Survey of Remote Sensing Image Fusion Assessment.Fire Control & Command Control, 2010, 35(2): 2-5 (in Chinese)
[8] 童明强.红外图像与可见光图像融合的研究: [硕士学位论文].天津: 天津理工大学, 2005 Tong Mingqiang.Research on Image Fusion of Infrared Image and Visible-Light Image.Master Degree Dissertation, Tianjing: Tianjing Insitute of Technology, 2005 (in Chinese)
相关文献:
1.朱亚辉, 彭国华.基于奇异值分解的图像融合效果综合评价研究[J]. 西北工业大学, 2013,31(1): 25-28