【摘要】：高光譜遙感是以遙感影像與光譜的合一為特征的新型遙感技術(shù),其光譜分辨率可高達(dá)納米級(jí),是20世紀(jì)80年代以來地球觀測(cè)技術(shù)所取得的重大技術(shù)突破之一,在現(xiàn)代軍事、礦物勘測(cè)、精確農(nóng)業(yè)以及環(huán)境監(jiān)控等領(lǐng)域有著廣泛的應(yīng)用。因此,研究高光譜數(shù)據(jù)的高效處理與解譯具有重要的理論意義和實(shí)際應(yīng)用價(jià)值。高光譜分類與目標(biāo)檢測(cè)是高光譜數(shù)據(jù)處理的主要內(nèi)容。高光譜數(shù)據(jù)提供了豐富的光譜信息,為人們研究地表物體的特性、進(jìn)行地物識(shí)別創(chuàng)造了條件。但是,這些海量的信息和特殊的數(shù)據(jù)結(jié)構(gòu)又給人們?cè)趫D像處理、信息分析、分類和檢測(cè)等方面提出了嚴(yán)峻的挑戰(zhàn),也就要求人們從多個(gè)方面去理解和揭示其物理特征及其變化。本論文在總結(jié)高光譜分類與目標(biāo)檢測(cè)現(xiàn)狀的基礎(chǔ)上,通過對(duì)高光譜數(shù)據(jù)自身特性的深入分析,挖掘高光譜數(shù)據(jù)的稀疏與低秩先驗(yàn)知識(shí),研究了聯(lián)合空譜信息的高光譜分類與目標(biāo)檢測(cè)技術(shù),并設(shè)計(jì)了相應(yīng)的高效算法。論文的主要工作和研究成果如下:(1)針對(duì)高光譜分類問題,提出了一種基于自適應(yīng)上下文信息的聯(lián)合稀疏表示分類方法。高光譜圖像不僅具有豐富的光譜信息,而且圖像中的每個(gè)像元還含有其自身的空間結(jié)構(gòu)。為了充分利用這種空間結(jié)構(gòu)信息,本文利用高階steering核函數(shù)來刻畫各個(gè)像元的局部空間結(jié)構(gòu),并將所得到的空間結(jié)構(gòu)與聯(lián)合稀疏表示模型融合得到最終的分類結(jié)果。結(jié)果表明,基于自適應(yīng)上下文信息的聯(lián)合稀疏表示分類方法可以有效描述地物的空間上下文信息,提高分類精度。(2)通過充分挖掘高光譜圖像的全局和局部空間信息,提出了一種基于低秩矩陣分解的高光譜圖像分類方法。首先采用一種基于圖的分割算法,將高光譜圖像分成多個(gè)勻質(zhì)區(qū)域。由同一勻質(zhì)區(qū)域內(nèi)的像元組成的矩陣(矩陣的列向量對(duì)應(yīng)像元的光譜向量)具有很強(qiáng)的列相關(guān)性,因此可將該二維矩陣分解成低秩矩陣和稀疏矩陣之和。然后以低秩矩陣中的列向量作為相應(yīng)像元的特征,利用概率支持向量機(jī)(Probabilistic support vector machine,PSVM)進(jìn)行分類。同時(shí)為了精細(xì)化分類結(jié)果,在PSVM中引入馬爾科夫隨機(jī)場(chǎng)正則化,保證高光譜圖像中的全局空間信息、局部空間信息和光譜信息有效結(jié)合。實(shí)驗(yàn)結(jié)果表明,該方法的分類結(jié)果優(yōu)于其他主流分類方法。(3)針對(duì)高光譜圖像異常檢測(cè)問題,提出了一種基于低秩和稀疏表示的高光譜異常檢測(cè)方法。該方法將高光譜圖像分解為背景和異常兩部分進(jìn)行分別建模,對(duì)于異常部分,根據(jù)異常像元在整個(gè)圖像中只占極少部分,本文將異常部分建模為一個(gè)列稀疏矩陣,并采用l2,1范數(shù)來刻畫。對(duì)于背景部分,由于高光譜圖像中的背景像元可認(rèn)為是位于多個(gè)子空間,采用低秩表示模型進(jìn)行建模,在背景字典下搜尋數(shù)據(jù)的最低秩表示,從而有效刻畫背景的全局結(jié)構(gòu)。為了能夠準(zhǔn)確表示每個(gè)像元光譜,刻畫像元的局部光譜信息,對(duì)其表示系數(shù)進(jìn)行稀疏約束。同時(shí),由于背景字典中的原子須要覆蓋所有背景地物種類且不能為異常像元,本文提出了一種新的背景字典構(gòu)造方法。針對(duì)所提模型,設(shè)計(jì)一個(gè)有效算法求解。在模擬和真實(shí)高光譜圖像上的實(shí)驗(yàn)表明,本文所提出的異常檢測(cè)方法能夠抑制背景像元突出異常像元,達(dá)到了較高的檢測(cè)準(zhǔn)確率。(4)針對(duì)高光譜視頻序列氣體檢測(cè)問題,提出了一種基于時(shí)空TV(Total Variation)正則化的目標(biāo)檢測(cè)方法。根據(jù)高光譜視頻序列中的氣體在空間維和時(shí)間維都具有連續(xù)性,在RPCA(Robust principal component analysis)模型中引入時(shí)空TV正則項(xiàng)刻畫氣體的空間連續(xù)性和時(shí)間連續(xù)性。同時(shí)為了充分利用高光譜視頻的光譜信息,采用主成分分析方法抽取每幀圖像中的主要特征,設(shè)計(jì)了一種多特征融合的氣體檢測(cè)方法。實(shí)驗(yàn)結(jié)果表明本文所提出的時(shí)空TV正則化模型能夠有效刻畫氣體的結(jié)構(gòu),提高檢測(cè)精度。
[Abstract]:Hyperspectral remote sensing is a new remote sensing technology characterized by the combination of remote sensing images and spectra. Its spectral resolution can reach nanometer level. It is one of the important breakthroughs in earth observation technology since the 1980s. It has been widely used in modern military, mineral exploration, precision agriculture and environmental monitoring. Hyperspectral classification and target detection are the main contents of hyperspectral data processing. Hyperspectral data provide abundant spectral information for people to study the characteristics of surface objects and identify them. Information and special data structure have posed a serious challenge to people in image processing, information analysis, classification and detection, which requires people to understand and reveal its physical characteristics and changes from many aspects. In-depth analysis of characteristics, mining sparse and low-rank prior knowledge of hyperspectral data, hyperspectral classification and target detection technology based on joint spatial spectrum information are studied, and corresponding efficient algorithms are designed. The hyperspectral image not only has abundant spectral information, but also each pixel in the image contains its own spatial structure. In order to make full use of this spatial structure information, this paper uses high-order steering kernel function to characterize the local spatial structure of each pixel, and the resulting spatial structure and structure. The results show that the combined sparse representation and classification method based on adaptive context information can effectively describe the spatial context information of objects and improve the classification accuracy. (2) A low rank moment based method is proposed by fully mining the global and local spatial information of hyperspectral images. A method of hyperspectral image classification based on matrix decomposition is presented. Firstly, a graph-based segmentation algorithm is used to divide hyperspectral images into homogeneous regions. Then, the column vectors of the low rank matrix are used as the feature of the corresponding pixels, and the probabilistic support vector machine (PSVM) is used to classify them. Experimental results show that the proposed method outperforms other mainstream classification methods. (3) To solve the problem of hyperspectral image anomaly detection, a hyperspectral anomaly detection method based on low rank and sparse representation is proposed. For the abnormal part, the abnormal part is modeled as a column sparse matrix and characterized by l2,1 norm. In order to represent the local spectral information of each pixel accurately, the representation coefficients of each pixel are sparsely constrained. At the same time, the atoms in the background dictionary must cover all species in the background and can not be abnormal images. An efficient algorithm is designed to solve the proposed model. Experiments on simulated and real hyperspectral images show that the proposed anomaly detection method can restrain background pixels from highlighting abnormal pixels and achieve a high detection accuracy. (4) For hyperspectral video sequences, the proposed method is effective. In this paper, a target detection method based on the regularization of space-time TV (Total Variation) is proposed for column gas detection. According to the continuity of gas in both space and time dimensions in hyperspectral video sequences, the space-time TV regularization term is introduced into the RPCA (Robust principal component analysis) model to characterize the space continuity and time continuity of gas. At the same time, in order to make full use of the spectral information of hyperspectral video, the principal component analysis (PCA) method is used to extract the main features of each image, and a multi-feature fusion gas detection method is designed.
【學(xué)位授予單位】：南京理工大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2016
【分類號(hào)】：TP751

【相似文獻(xiàn)】

相關(guān)期刊論文 前3條

1 高曉惠;相里斌;衛(wèi)俊霞;魏儒義;于濤;;基于空間特征的光譜分類算法研究[J];光譜學(xué)與光譜分析;2010年10期

2 王增柱,劉同懷,黃魯;基于光譜分類的超光譜數(shù)據(jù)壓縮方法[J];遙感技術(shù)與應(yīng)用;2001年03期

3 ;[J];;年期

相關(guān)博士學(xué)位論文 前1條

1 徐洋;稀疏與低秩先驗(yàn)下的高光譜分類與檢測(cè)方法[D];南京理工大學(xué);2016年

相關(guān)碩士學(xué)位論文 前2條

1 葉虎平;黃東海海洋光譜分類方法研究[D];國(guó)家海洋技術(shù)中心;2009年

2 羅磊;基于模糊模式識(shí)別的光譜分類研究[D];鄭州大學(xué);2006年

，

本文編號(hào)：2182670