本書共計11章,第1章對合成孔徑雷達(SAR)目標識別進行了概述;第2章介紹了基于局部保持特性和混合高斯分布的SAR目標識別;第3章介紹了基于局部保持特性和Gamma分布的SAR目標識別;第4章介紹了基于結(jié)構(gòu)保持投影的SAR目標識別;第5章介紹了基于類別稀疏表示的SAR目標識別;第6章介紹了基于乘性稀疏表示和Gamma分布的SAR目標識別;第7章介紹了基于判別統(tǒng)計字典學習的SAR目標識別;第8章介紹了于Dempster-Shafer證據(jù)理論融合多稀疏描述和樣本統(tǒng)計特性的SAR目標識別;第9章介紹了基于Dempster-Shafer證據(jù)理論和稀疏表示的SAR目標識別;第10章介紹了基于兩階段稀疏結(jié)構(gòu)表示的SAR目標識別;第11章探討了未來合成孔徑雷達目標識別可能的發(fā)展方向。
劉明,工學博士,副教授,碩士生導師。2009年獲西安電子科技大學信息對抗技術(shù)專業(yè)工學學士學位,2015年獲西安電子科技大學模式識別與智能系統(tǒng)專業(yè)工學博士學位。2019年-2020年為加拿大McMaster University訪學學者。主要研究方向為:目標檢測與目標識別。入選陜西省科協(xié)青年人才托舉計劃,獲國際無線電科學聯(lián)盟(URSI)"青年科學家”獎,獲陜西省計算機學會"計算機領(lǐng)域優(yōu)秀青年專家”稱號。主持和參與了包括國家自然科學基金、國家重大基礎(chǔ)研究計劃、裝備預先研究、陜西省自然科學基金等10余項國家級和省部級科研項目。發(fā)表學術(shù)論文60余篇,授權(quán)國家發(fā)明專利10項(部分已轉(zhuǎn)化)。
第1 章 緒論························································································1
1.1 研究背景及研究意義··································································1
1.2 國內(nèi)外研究現(xiàn)狀········································································3
1.3 本書內(nèi)容介紹········································································.10
第2 章 基于局部保持特性和混合高斯分布的SAR 圖像目標識別··················.14
2.1 算法概述··············································································.14
2.2 局部保持投影算法··································································.15
2.3 基于LPP-GMD 算法的SAR 圖像目標識別···································.16
2.3.1 基于混合高斯分布的似然函數(shù)建!ぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁ.17
2.3.2 基于局部保持特性的先驗函數(shù)建!ぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁ.17
2.3.3 參數(shù)估計·····································································.18
2.4 試驗結(jié)果與分析·····································································.22
2.5 本章小結(jié)··············································································.26
第3 章 基于局部保持特性和Gamma 分布的SAR 圖像目標識別··················.27
3.1 算法概述··············································································.27
3.2 SAR 圖像的乘性相干斑模型······················································.28
3.3 基于LPP-Gamma 算法的SAR 圖像目標識別·································.29
3.3.1 基于Gamma 分布構(gòu)建似然函數(shù)········································.29
3.3.2 基于局部保持特性構(gòu)建先驗函數(shù)·······································.30
3.3.3 參數(shù)估計·····································································.33
3.4 試驗結(jié)果與分析·····································································.37
3.4.1 SAR 圖像目標識別結(jié)果··················································.37
3.4.2 修正相似度矩陣的有效性驗證··········································.39
3.5 本章小結(jié)··············································································.41
第4 章 基于結(jié)構(gòu)保持投影的SAR 圖像目標識別·······································.42
4.1 算法概述··············································································.42
4.2 基于CDSPP 算法的SAR 圖像目標識別·······································.43
4.2.1 CDSPP 算法·································································.43
4.2.2 差異度矩陣分析····························································.45
4.3 試驗結(jié)果與分析·····································································.49
4.3.1 目標的類別識別····························································.51
4.3.2 目標的型號識別····························································.53
4.3.3 構(gòu)建差異度矩陣的優(yōu)勢···················································.57
4.4 本章小結(jié)··············································································.59
第5 章 基于類別稀疏表示的SAR 圖像目標識別·······································.60
5.1 算法概述··············································································.60
5.2 SAR 圖像的稀疏表示模型·························································.61
5.3 SAR 圖像的類別稀疏表示模型···················································.62
5.3.1 方位角敏感特性····························································.62
5.3.2 測試樣本建!ぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁぁ.64
5.3.3 稀疏向量求解·······························································.66
5.4 基于LSR 算法的SAR 圖像目標識別···········································.67
5.5 試驗結(jié)果與分析·····································································.70
5.5.1 目標的類別識別····························································.70
5.5.2 目標的型號識別····························································.72
5.6 本章小結(jié)··············································································.76
第6 章 基于乘性稀疏表示和Gamma 分布的SAR 圖像目標識別··················.77
6.1 算法概述··············································································.77
6.2 乘性稀疏表示算法··································································.78
6.3 試驗結(jié)果與分析·····································································.80
6.3.1 目標的類別識別····························································.81
6.3.2 目標的型號識別····························································.82
6.4 本章小結(jié)··············································································.88
第7 章 基于判別統(tǒng)計字典學習的SAR 圖像目標識別·································.89
7.1 算法概述··············································································.89
7.2 基于判別統(tǒng)計字典學習(DSDL)的SAR 圖像目標識別··················.90
7.2.1 統(tǒng)計字典學習(SDL)算法·············································.90
7.2.2 融入判別因子字典·························································.93
7.2.3 算法的計算復雜度分析···················································.94
7.3 試驗結(jié)果與分析·····································································.96
7.3.1 目標的類別識別····························································.97
7.3.2 目標的型號識別····························································.98
7.4 本章小結(jié)··············································································103
第8 章 基于Dempster-Shafer 證據(jù)理論融合多稀疏表示和樣本統(tǒng)計特性的SAR
圖像目標識別·········································································105
8.1 算法概述··············································································105
8.2 Dempster-Shafer 證據(jù)理論·························································106
8.3 基于Dempster-Shafer 證據(jù)理論的融合算法···································107
8.3.1 SAR 圖像的多稀疏表示················································.107
8.3.2 基本概率分配函數(shù)的推導··············································.113
8.4 試驗結(jié)果與分析·····································································117
8.5 本章小結(jié)··············································································119
第9 章 基于Dempster-Shafer 證據(jù)理論和稀疏表示的SAR 圖像目標識別······120
9.1 算法概述··············································································120
9.2 基于Dempster-Shafer 證據(jù)理論的融合算法···································121
9.2.1 構(gòu)建基于稀疏表示的基本概率分配函數(shù)····························.121
9.2.2 融合算法···································································.123
9.3 試驗結(jié)果與分析·····································································125
9.3.1 目標的類別識別··························································.126
9.3.2 目標的型號識別··························································.128
9.4 本章小結(jié)··············································································131
第10 章 基于兩階段稀疏結(jié)構(gòu)表示的SAR 圖像目標識別····························132
10.1 算法概述·············································································132
10.2 基于兩階段稀疏結(jié)構(gòu)表示(TSSR)的算法··································133
10.2.1 第一階段(訓練階段)的結(jié)構(gòu)保持································.133
10.2.2 第二階段(測試階段)的結(jié)構(gòu)保持································.135
10.3 試驗結(jié)果與分析····································································140
10.3.1 目標的類別識別·························································.141
10.3.2 目標的型號識別·························································.145
10.4 本章小結(jié)·············································································150
第11 章 總結(jié)與展望···········································································151
11.1 全書總結(jié)·············································································151
11.2 工作展望·············································································153
參考文獻···························································································155