首页期刊简介编委会投稿启事审稿流程读者订阅广告服务联系我们English
引用本文
  •    [点击复制]
  •    [点击复制]
【打印本页】 【下载PDF全文】 查看/发表评论下载PDF阅读器关闭

←前一篇|后一篇→

过刊浏览    高级检索

本文已被:浏览 216次   下载 0 本文二维码信息
码上扫一扫!
决策可靠性分析及在SAR图像目标识别中的应用
靳黎忠,陈俊杰,彭新光
0
(1.太原理工大学 信息与计算机学院,太原 030024;2.太原科技大学 应用科学学院,太原 030024)
摘要:
决策融合是提高合成孔径雷达(Synthetic Aperture Radar,SAR)目标识别性能的重要手段,然而,可靠性较弱的决策往往会导致最终决策融合的效果变差。将可靠性分析引入基于决策融合的SAR目标识别方法中,分别计算各个决策的可靠性系数并选取可靠性的决策参与最终的决策融合。为了验证方法的有效性,分别将提出的可靠性分析应用于多特征决策融合以及多分类器决策融合并基于MSTAR(Moving and Stationary Target Acquisition and Recognition)数据集进行了目标识别实验。在基于主成分分析、线性鉴别分析和非负矩阵分解三种特征进行多特征决策融合的条件下,所提方法和直接进行决策融合的方法的识别率分别为97.47%和96.50%。在基于K近邻、支持向量机和稀疏表示分类器的多分类器决策融合中,所提方法和直接进行决策融合的方法的识别率分别为97.10%和96.28%。实验结果充分证明了所提方法的有效性。
关键词:  合成孔径雷达  目标识别  决策融合  可靠性分析
DOI:
基金项目:国家自然科学基金资助项目(61672374)
Reliability analysis for decision fusion and its application in target recognition of SAR images
JIN Lizhong,CHEN Junjie,PENG Xinguang
(1.College of Information and Computer,Taiyuan University of Technology,Taiyuan 030024,China;2.School of Applied Science,Taiyuan University of Science and Technology,Taiyuan 030024,China)
Abstract:
Decision fusion is an effective way to improve synthetic aperture radar(SAR) target recognition performance.However,the decisions with low reliabilities will impair the fused performance to some extent.Therefore,this paper brings reliability analysis into decision fusion for SAR target recognition.The reliability levels of individual decisions are calculated and only those with high reliability levels are used in the final decision fusion.To validate the effectiveness of the proposed method,the proposed strategy is applied in multi-feature decision fusion and multi-classifier decision and target recognition experiments are conducted on Moving and Stationary Target Acquisition and Recognition(MSTAR) dataset.Principle component analysis(PCA),liner discriminant analysis(LDA),and non-negative matrix factorization(NMF) are used for feature extraction in the multi-feature decision fusion.The proposed method and the direct decision fusion achieve the recognition accuracies of 97.47% and 96.50%,respectively.K-nearest neighbor(KNN),support vector machine(SVM),and sparse representation-based classification(SRC) are used as the classifiers in the multi-classifier decision fusion.The proposed method and the direct decision fusion achieve the recognition accuracies of 97.10% and 96.28%,respectively.The experimental results demonstrate the effectiveness of the proposed method.
Key words:  synthetic aperture radar(SAR)  target recognition  decision fusion  reliability analysis
安全联盟站长平台