本文選題：高光譜圖像　切入點(diǎn)：遙感　出處：《南京航空航天大學(xué)》2017年碩士論文

【摘要】：高光譜遙感技術(shù)的快速發(fā)展,提高了拍攝場(chǎng)景的信息豐富度,為遙感圖像中的異常(小目標(biāo))檢測(cè)開辟了新的途徑,賦予了其更重要的實(shí)際意義。為了構(gòu)建一個(gè)完整的高光譜遙感圖像異常檢測(cè)系統(tǒng),實(shí)現(xiàn)異常成分的自動(dòng)定位,需要對(duì)系統(tǒng)中涉及的波段選擇、特征提取、異常檢測(cè)等方法進(jìn)行研究。本文的主要工作如下:首先,探索了一種基于子空間中主成分最優(yōu)線性預(yù)測(cè)的波段選擇方法。采用改進(jìn)相關(guān)性度量的譜聚類方法將高光譜波段劃分為不同的子空間,并對(duì)各子空間中的波段進(jìn)行主成分分析(Principal Component Analysis,PCA),選擇主要分量作為重構(gòu)目標(biāo);以子空間追蹤法為搜索策略,從各子空間中選擇數(shù)個(gè)波段對(duì)其重構(gòu)目標(biāo)進(jìn)行聯(lián)合最優(yōu)線性預(yù)測(cè);合并各子空間中的所選波段得到最佳波段子集。實(shí)驗(yàn)結(jié)果表明,該方法選擇的波段子集可以較完整地重構(gòu)原始數(shù)據(jù),與原始數(shù)據(jù)以及自適應(yīng)波段選擇(Auto Band Selection,ABS)方法、線性預(yù)測(cè)(Linear Prediction,LP)方法、最大方差主成分分析(Maximum-variance Principal Component Analysis,MVPCA)方法、自相關(guān)矩陣波段選擇(Auto Correlation Matrix-based Band Selection,ACMBS)方法得到的波段子集相比,其波段子集具有更好的異常檢測(cè)性能。然后,討論了一種基于模糊監(jiān)督與流形結(jié)構(gòu)保持的特征提取方法。采用RX方法粗略計(jì)算像元隸屬于異常的可能性,并將其作為模糊監(jiān)督信息;在模糊監(jiān)督信息的基礎(chǔ)上,分別建立同類局部、非同類非局部以及全局的流形結(jié)構(gòu)近鄰圖,并結(jié)合這3種近鄰圖構(gòu)建低維映射的目標(biāo)函數(shù);為了使該方法能夠處理新像元并適應(yīng)非線性結(jié)構(gòu)數(shù)據(jù),給出了其線性化與核化方法,以獲得線性與非線性投影矢量。實(shí)驗(yàn)結(jié)果表明,與主成分分析方法、核主成分分析(Kernel Principal Component Analysis,KPCA)方法、局部線性嵌入(Locally Linear Embedding,LLE)方法、局部保持投影(Locality Preserving Projection,LPP)方法相比,該方法的特征數(shù)據(jù)擁有更高的異常像元顯著性,能夠獲得更佳的異常檢測(cè)結(jié)果。其次,研究了一種基于背景聚類與加權(quán)迭代RX的異常檢測(cè)方法。采用改進(jìn)密度峰值快速搜索算法將高光譜像元?jiǎng)澐譃椴煌谋尘邦惾?在此基礎(chǔ)上,結(jié)合3種方法對(duì)RX檢測(cè)窗中的外窗像元進(jìn)行加權(quán),以獲得更準(zhǔn)確的背景統(tǒng)計(jì)模型:依據(jù)外窗像元與各背景類群的馬氏距離加權(quán),以減小外窗中異常像元的權(quán)重;根據(jù)各背景類群對(duì)待檢測(cè)像元的貢獻(xiàn)度加權(quán),以降低外窗中相異類群背景像元的影響;利用初次檢測(cè)結(jié)果加權(quán)后再迭代檢測(cè),達(dá)到進(jìn)一步純化外窗中背景的目的。實(shí)驗(yàn)結(jié)果表明,與常規(guī)RX方法、分塊自適應(yīng)異常點(diǎn)計(jì)算(Blocked Adaptive Computationally Efficient Outlier Nominators,BACON)方法,異常加權(quán)RX(Weighted Anomaly RX,WARX)方法、概率異常檢測(cè)(Probabilistic Anomaly Detector,PAD)方法相比,該方法不僅可以明顯提高RX方法對(duì)檢測(cè)窗口尺寸的魯棒性,而且能夠獲得更高的檢測(cè)精度。再次,提出了一種基于自適應(yīng)參數(shù)支持向量機(jī)(Support Vector Machine,SVM)的異常檢測(cè)方法。通過無監(jiān)督檢測(cè)方法對(duì)異常像元進(jìn)行快速、粗略定位,并將該定位結(jié)果作為后驗(yàn)信息輸入到支持向量機(jī)中;依據(jù)后驗(yàn)信息與核空間散度準(zhǔn)則自適應(yīng)確定支持向量機(jī)中核函數(shù)的參數(shù),并使用該支持向量機(jī)在核空間中尋找分離異常和背景的最佳超平面;利用該超平面將像元重新分類為背景和異常,并且迭代上述操作,得到穩(wěn)定的異常檢測(cè)結(jié)果。實(shí)驗(yàn)結(jié)果表明,與常規(guī)RX方法、核RX(Kernel RX,KRX)方法、支持向量數(shù)據(jù)描述(Support Vector Data Description,SVDD)方法相比,該方法可以更有效、精確地檢測(cè)出高光譜遙感圖像中的異常成分。最后,提出了一種基于蜂群優(yōu)化投影尋蹤(Projection Pursuit,PP)與加權(quán)K最近鄰(Weighted K-nearest Neighbor,WKNN)的異常檢測(cè)方法。結(jié)合鄰域像元聯(lián)合定義的峰度與偏度為投影指標(biāo),以MABC為尋優(yōu)方法,使用投影尋蹤從高光譜圖像中逐次獲取投影圖像,再根據(jù)其直方圖提取異常像元;在初檢結(jié)果的基礎(chǔ)上,提取包含像元判別信息與主要結(jié)構(gòu)的特征,結(jié)合WKNN方法對(duì)初檢結(jié)果進(jìn)行提純。實(shí)驗(yàn)結(jié)果表明,與RX方法、獨(dú)立分量分析(Independent Component Analysis,ICA)方法以及混沌粒子群優(yōu)化(Chaotic Particle Swarm Optimization,CPSO)投影尋蹤方法相比,該方法不但可以獲得虛警率更低的異常檢測(cè)結(jié)果,而且具有更快的運(yùn)算速度。
[Abstract]:The rapid development of hyperspectral remote sensing technology, improve the shooting scene information richness, abnormal in remote sensing image (target) provides a new way to detect, given its important practical significance. In order to build a complete hyperspectral remote sensing image anomaly detection system, automatic positioning and implementation of abnormal components. To extract the features of the selection, involving the band system, studied the anomaly detection method. The main work of this paper is as follows: firstly, to explore a method to select the principal component subspace optimal linear prediction based on the improved correlation metric band. The spectral clustering method of high spectral band is divided into different sub spaces. In the space of the bands in the principal component analysis (Principal Component, Analysis, PCA), select the main component as the target for reconstruction; search strategy search for subspace tracking method, from the air In the choose a number of bands are combined to reconstruct the optimal linear prediction; with each subspace in the selected band get the best band subsets. The experimental results show that this method can select band subset of a complete reconstruction of the original data with the original data and adaptive band selection (Auto Band Selection, ABS) method of linear prediction (Linear Prediction LP) method, principal component analysis with varimax (Maximum-variance Principal Component Analysis, MVPCA) method, choose the autocorrelation matrix band (Auto Correlation Matrix-based Band Selection, ACMBS) compared with the band subset obtained, the subset of the performance anomaly detection has better. Then, is discussed. A fuzzy feature extraction method based on manifold structure and maintain supervision. By the method of RX pixels belonging to rough calculation the possibility of abnormal, and as a fuzzy Based on fuzzy information supervision; supervision information, establish similar local manifold structure, nearest neighbor graph of non similar non local and global, and combined with the objective function of the 3 nearest neighbor graph based low dimensional mapping; in order to make the method can handle the new pixel and adapt to the nonlinear structure of the data, given its linearization and nuclear method to obtain the linear and nonlinear projection vector. The experimental results show that the analysis method and the principal component, kernel principal component analysis (Kernel Principal Component Analysis, KPCA), locally linear embedding (Locally Linear Embedding, LLE) method, locality preserving projection (Locality Preserving, Projection, LPP) compared with other methods, the method of data characteristics have a higher significant pixel anomaly, anomaly detection can obtain better results. Secondly, the anomaly detection method is studied based on the background of clustering and weighted iterative RX mining. With the improvement of the peak density of fast search algorithm for hyperspectral pixel is divided into groups of different background, on the basis of this, 3 methods are weighted window RX pixels in the detection window, to obtain a statistical background model is more accurate: Based on Mahalanobis distance weighted window pixel and the background of the group, to reduce weight the abnormal pixels outside the window; according to the weighted degree for each background pixel groups to be detected with, in order to reduce the influence of external window in different groups of background pixels; using the initial test results after weighted iterative detection, to achieve further purification in the background window. The experimental results show that with the conventional RX method, block adaptive anomaly (Blocked Adaptive Computationally Efficient calculation of Outlier Nominators, BACON) method, weighted RX (Weighted Anomaly RX anomaly, WARX) method, the probability of anomaly detection (Probabilistic Anomaly Detector, P AD) compared with other methods, this method not only can improve the robustness of RX method for detecting the size of the window, and can obtain higher detection accuracy. Thirdly, we propose a support vector machine based on adaptive parameters (Support Vector, Machine, SVM) anomaly detection method. The detection method of abnormal fast unsupervised pixel rough, positioning, and the positioning results as a posteriori information input to the support vector machine; based on a posteriori information and spatial divergence criterion to adaptively determine the nuclear parameters of kernel function for support vector machine, and use the support vector machine to find the best hyperplane separating anomaly and background in the kernel space; the super plane the pixel is reclassified as background and anomaly, and the iterative operation, get the results of anomaly detection stability. Experimental results show that with the conventional RX method, RX (Kernel RX, KRX) method, support vector Data description (Support Vector Data Description, SVDD) compared with other methods, this method can effectively and accurately detect abnormal components of hyperspectral remote sensing images. Finally, proposes a bee colony optimization based on projection pursuit (Projection Pursuit, PP) and the weighted K nearest neighbor (Weighted K-nearest, Neighbor, WKNN) anomaly detection methods. The combination of kurtosis and skewness of neighborhood pixels defined as joint projection index, with MABC as the optimization method, using projection pursuit from hyperspectral image to obtain successive projection images according to the histogram extraction abnormal pixel; based on initial inspection results, including extraction characteristics of pixels discriminant information and main structure, combined with WKNN method for purification of the initial inspection results. The experimental results show that, with the RX method, independent component analysis (Independent Component, Analysis, ICA) method and chaotic particle swarm optimization (Chaotic Particle Compared with the Swarm Optimization, CPSO) projection pursuit method, this method can not only obtain the abnormal detection results with lower false alarm rate, but also have faster operation speed.

【學(xué)位授予單位】：南京航空航天大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP751

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 張鵬強(qiáng);譚熊;余旭初;魏祥坡;薛志祥;;基于核半監(jiān)督判別分析的高光譜影像特征提取[J];測(cè)繪科學(xué)技術(shù)學(xué)報(bào);2016年03期

2 方帥;瞿成佳;楊學(xué)志;劉永進(jìn);;組合因子最優(yōu)的線性預(yù)測(cè)波段選擇[J];中國圖象圖形學(xué)報(bào);2016年02期

3 王立國;趙亮;劉丹鳳;;基于人工蜂群算法高光譜圖像波段選擇[J];哈爾濱工業(yè)大學(xué)學(xué)報(bào);2015年11期

4 趙春暉;崔曉辰;;基于全局和局部流形結(jié)構(gòu)的高光譜圖像特征提取算法[J];沈陽大學(xué)學(xué)報(bào)(自然科學(xué)版);2015年04期

5 秦方普;張愛武;王書民;孟憲剛;胡少興;孫衛(wèi)東;;基于譜聚類與類間可分性因子的高光譜波段選擇[J];光譜學(xué)與光譜分析;2015年05期

6 許明明;張良培;杜博;張樂飛;;基于類別可分性的高光譜圖像波段選擇[J];計(jì)算機(jī)科學(xué);2015年04期

7 吳一全;曹照清;陶飛翔;;結(jié)合多尺度幾何分析和KICA的遙感圖像變化檢測(cè)[J];遙感學(xué)報(bào);2015年01期

8 吳一全;殷駿;戴一冕;袁永明;;基于蜂群優(yōu)化多核支持向量機(jī)的淡水魚種類識(shí)別[J];農(nóng)業(yè)工程學(xué)報(bào);2014年16期

9 趙春暉;李曉慧;王玉磊;;高光譜圖像異常目標(biāo)檢測(cè)研究進(jìn)展[J];電子測(cè)量與儀器學(xué)報(bào);2014年08期

10 張海濤;王鶴橋;孟祥羽;武文波;;基于類對(duì)可分和灰色決策的高光譜波段選擇方法[J];計(jì)算機(jī)科學(xué);2014年06期

相關(guān)碩士學(xué)位論文 前2條

1 王立志;基于流形學(xué)習(xí)的高光譜圖像降維與分類研究[D];重慶大學(xué);2012年

2 王超;基于流形學(xué)習(xí)的有監(jiān)督降維方法研究[D];中國科學(xué)技術(shù)大學(xué);2009年

，

本文編號(hào)：1706860