Representative band selection for hyperspectral image classification

High dimensional curse for hyperspectral images is one major challenge in image classification. In this work, we introduce a novel spectral band selection method by representative band mining. In the proposed method, the distance between two spectral bands is measured by using disjoint information. For band selection, all spectral bands are first grouped into clusters, and representative bands are selected from these clusters. Different from existing clustering-based band selection methods which select bands from each cluster individually, the proposed method aims to select representative bands simultaneously by exploring the relationship among all band clusters. The optimal representative band selection is based on the criteria of minimizing the distance inside each cluster and maximizing the distance among different representative bands. These selected bands can be further applied in hyperspectral image classification. Experiments are conducted on the 92AV3C Indian Pine data set. Experimental results show that the disjoint information-based spectral band distance measure is effective and the proposed representative band selection approach outperforms state-of-the-art methods for high dimensional image classification.     

利用独立成分分析的高光谱图像波段选择方法

提出一种适合目标探测的基于独立成分分析(ICA)的高光谱图像波段选择方法。首先进行"虚拟维"(VD)估计以确定重要独立成分个数,同时对FastICA生成的独立成分排序,选择排序靠前的几个独立成分作为重要独立成分;再根据波段对重要独立成分的平均贡献量对波段排序;最后使用光谱相似性度量去除排序后的冗余波段,保证了最终波段子集含有较多的目标信息。对AVIRIS获取的两幅真实高光谱图像进行了目标探测实验,结果表明,文中方法优于另外两种基于二阶统计特性的波段选择方法,其选出的波段分别占据全部波段的12%和3%,目标探测算子自适应余弦估计(ACE)和自适应匹配滤波(AMF)其上的探测率较全波段分别提高了30%和15%。     

基于子空间中主成分最优线性预测的高光谱波段选择

针对高光谱遥感图像的异常检测问题,为了使高光谱降维数据能更完整地保留其光谱信息,提出了基于子空间中主成分最优线性预测的波段选择方法.采用改进相关性度量的谱聚类方法将高光谱波段划分为不同的子空间,并对各子空间中的波段进行主成分分析(PCA),选择主要分量作为重构目标;以子空间追踪法为搜索策略,从各子空间中选择数个波段对其重构目标进行联合最优线性预测;合并各子空间中的所选波段得到最佳波段子集.实验结果表明,该方法选择的波段子集可以较完整地重构原始数据,与原始数据以及自适应波段选择(ABS)方法、线性预测(LP)方法、最大方差主成分分析(MVPCA)方法、自相关矩阵波段选择(ACMBS)方法、组合因子最优波段选择(OCFBS)方法得到的波段子集相比,其波段子集具有更好的异常检测性能.     

Constrained band selection for hyperspectral imagery

Constrained energy minimization (CEM) has shown effective in hyperspectral target detection. It linearly constrains a desired target signature while minimizing interfering effects caused by other unknown signatures. This paper explores this idea for band selection and develops a new approach to band selection, referred to as constrained band selection (CBS) for hyperspectral imagery. It interprets a band image as a desired target signature vector while considering other band images as unknown signature vectors. As a result, the proposed CBS using the concept of the CEM to linearly constrain a band image, while also minimizing band correlation or dependence provided by other band images, is referred to as CEM-CBS. Four different criteria referred to as Band Correlation Minimization (BCM), Band Correlation Constraint (BCC), Band Dependence Constraint (BDC), and Band Dependence Minimization (BDM) are derived for CEM-CBS.. Since dimensionality resulting from conversion of a band image to a vector may be huge, the CEM-CBS is further reinterpreted as linearly constrained minimum variance (LCMV)-based CBS by constraining a band image as a matrix where the same four criteria, BCM, BCC, BDC, and BDM, can be also used for LCMV-CBS. In order to determine the number of bands required to select p, a recently developed concept, called virtual dimensionality, is used to estimate the p. Once the p is determined, a set of p desired bands can be selected by the CEM/LCMV-CBS. Finally, experiments are conducted to substantiate the proposed CEM/LCMV-CBS four criteria, BCM, BCC, BDC, and BDM, in comparison with variance-based band selection, information divergence-based band selection, and uniform band selection.     

基于分类别PCA散度的高光谱图像分类波段选择

波段选择是去除高光谱图象段间冗余,实现降维的有效方法。该文提出了一种新的基于分类别主成分分析(PCA)散度的波段选择方法。即首先对训练集各类样本分别进行PCA变换去相关并计算散度,接着分析相应PCA变换系数获得对各类样本分类都重要的原始波段,在综合考虑波段的相关度,散度和子集规模的基础上获得最终选择波段。复杂度分析表明该方法较局部寻优的前向搜索计算量大为降低,提高了效率,并用高光谱遥感图象的分类实验进行了验证。     

一种结合空谱聚类的高光谱图像快速压缩算法

对高光谱图像进行快速压缩已经成为了高光谱遥感领域的研究热点.针对现有的高光谱图像数据量大和压缩所需运算量大的问题,提出了一种基于频段聚类+主成分分析(PCA)与空间分类相结合的高光谱图像快速压缩算法.首先利用最大相关度频段聚类算法(MCBC)将频段聚类,接着将每一类频段用PCA压缩,然后将压缩后的图像利用聚类信号子空间投影(CSSP)算法进行图像分类,最后在每一类内利用LBG(Linde Buzo Gray)算法通过矢量量化快速完成高光谱图像的编码.在不同的压缩比下进行实验,结果表明提出的高光谱图像压缩算法能在保证良好的图像恢复质量的前提下,大幅度降低运算复杂度,实现高光谱图像的快速压缩.     

基于自适应波段聚类主成分分析和反向传播神经网络的高光谱图像压缩

高光谱遥感图像具有丰富的光谱信息,数据量大。为了能够有效地利用高光谱图像数据,促进高光谱遥感技术的发展,该文提出一种基于自适应波段聚类主成分分析(PCA)与反向传播(BP)神经网络相结合的高光谱图像压缩算法。算法利用近邻传播(AP)聚类算法对波段进行自适应聚类,对聚类后的各个分组分别进行PCA运算,最后利用BP神经网络对所有主成分进行编码压缩。该文的创新点在于BP神经网络压缩图像时,在训练步骤过程中,误差反向传播是用原图与输出作差值,再反向调整各层的权值、阈值。对高光谱图像进行波段聚类,不仅能够有效地利用谱间相关性,提高压缩性能,还可以降低PCA的运算量。实验结果表明,该文算法与其它现有算法比较,在相同压缩比下,其光谱角更小,信噪比更高。     

基于谱间去相关模型的迭代硬阈值重构算法

根据高光谱图像的特点,提出一种基于谱间去相关模型的迭代硬阈值重构算法。根据高光谱图像序列的相邻图像之间具有很强的相关性,在迭代硬阈值重构算法中建立谱间去相关模型,除去重构图像观测数据中谱间相关的观测数据,去相关后的图像的观测数据更加稀疏,重构性能更高。实验结果表明,在相同观测数目下,本算法与迭代硬阈值重构算法相比,有效提高了图像的重构质量。     

Hyperion高光谱影像波段选择方法比较研究

Hyperion影像的光谱分辨率高,数据体积庞大,而且相邻波段之间的相关性强,信息冗余度较高, 给数据处理与解译带来了很多问题。鉴于此,提出了通过将分段主成分分析和波段指数相结合来开展波段选择与降维研究的思想。 同时采用自适应波段选择法、波段指数法和主成分分析累计贡献率方法进行了波段选择方法的对比研究;对4种波段选择方法所得到的结 果进行了最佳波段组合、地物可分性和图像变换比较分析。实验结果表明,分段主成分分析与波段指数综合方法可以有效抑制由于全局变换造成局部重要光谱被滤除的现象 ,同时还可兼顾自适应分区后各子区间及区间内波段之间的相关性,有效降低高光谱数据的维度。由此可见,该方法的波段选择效 果优于传统的自适应波段选择方法、波段指数法以及主成分分析累计贡献率方法。     

Band Selection in Multispectral Images by Minimization of Dependent Information

In this paper, a band selection technique for hyperspectral image data is proposed. Supervised feature extraction techniques allow a reduction of the dimensionality to extract relevant features through a labeled training set. This implies an analysis of the existing class distributions, which usually means, in the case of hyperspectral imaging, a large number of samples, making the labeling process difficult. A possible alternative could be the use of information measures, which are the basis of the proposed method. The present approach basically behaves as an unsupervised feature selection criterion, to obtain the relevant spectral bands from a set of sample images. The relations of information content between spectral bands are analyzed, leading to the proposed technique based on the minimization of the dependent information between spectral bands, while trying to maximize the conditional entropies of the selected bands     

Spectral band selection for visible-near infrared remote sensing:spectral-spatial resolution tradeoffs

Recent technological developments permit improved instrumentation for surveillance and resource monitoring, but tradeoffs of spectral resolution and number of spectral bands versus spatial resolution and measurement precision must be considered. A band selection procedure is applied to high spectral resolution (0.01 μ/m) aircraft sensor imagery representing the visible and near-infrared wavelengths (0.4-2.5 μm). Approximately 30-40 spectral bands characterize virtually all the information (variability) in the data, with the precise number depending on issues of data interpretation. This suggests that lower spectral resolution and higher spatial resolution are preferable to the reverse. Further study is needed to evaluate the significance of spectral bands having very low amplitude variability     

Independent component analysis-based dimensionality reduction with applications in hyperspectral image analysis

In hyperspectral image analysis, the principal components analysis (PCA) and the maximum noise fraction (MNF) are most commonly used techniques for dimensionality reduction (DR), referred to as PCA-DR and MNF-DR, respectively. The criteria used by the PCA-DR and the MNF-DR are data variance and signal-to-noise ratio (SNR) which are designed to measure data second-order statistics. This paper presents an independent component analysis (ICA) approach to DR, to be called ICA-DR which uses mutual information as a criterion to measure data statistical independency that exceeds second-order statistics. As a result, the ICA-DR can capture information that cannot be retained or preserved by second-order statistics-based DR techniques. In order for the ICA-DR to perform effectively, the virtual dimensionality (VD) is introduced to estimate number of dimensions needed to be retained as opposed to the energy percentage that has been used by the PCA-DR and MNF-DR to determine energies contributed by signal sources and noise. Since there is no prioritization among components generated by the ICA-DR due to the use of random initial projection vectors, we further develop criteria and algorithms to measure the significance of information contained in each of ICA-generated components for component prioritization. Finally, a comparative study and analysis is conducted among the three DR techniques, PCA-DR, MNF-DR, and ICA-DR in two applications, endmember extraction and data compression where the proposed ICA-DR has been shown to provide advantages over the PCA-DR and MNF-DR.     

波段最大筛选法及其在高光谱目标探测中的应用

提出了一种面向目标探测的高光谱图像波段选择方法—波段最大筛选法(MBS),它将每个波段的图像看成一条波段向量,以两个最不相似的波段作为初始波段,每次从剩余波段中选取一个和已选波段最不相似的波段,通过对波段相似性阈值的合理调节,保证了目标探测算子在所选波段上探测效果最佳。为了验证MBS的有效性,对机载可见光/红外成像光谱仪(AVIRIS)获取的两幅真实高光谱图像进行了实验,结果表明,MBS选出的波段分别占据全部波段的15%和9%,从而使目标探测算子ACE和AMF在其上的探测性能有了明显改善。     

一种基于改进子空间划分的波段选择方法

高光谱图像具有光谱分辨率高、波段连续、数据量大、图谱合一等特点。然而较高的光谱分辨率会造成波段间相关性强,信息冗余多。所以如何从数百个高光谱波段中选出有利于识别或分类的波段组合成为了高光谱应用需要解决的问题。文章针对相邻波段间相关性较大的特点,提出一种改进的对波段相关矩阵进行全局搜索的子空间划分的波段选择方法。该方法克服了传统只利用相关向量对波段进行划分的缺陷,利用整个相关矩阵进行全局搜索划分,再在划分后的子空间内进行波段选择,从而降低了波段之间的相关性。文章最后使用上述方法对AVIRIS数据进行波段选择,并通过SVM方法对其进行地物分类,结果表明该方法较不进行子空间划分的波段选择方法有较高的分类精度。     

一种基于主成分分析的高光谱图像波段选择算法

提出了一种基于主成分分析的高光谱图像波段选择算法。该算法把每个波段被映射到主成分的信息量的大小作为是否被选择的指标，因此，可以保证选择的波段包含原始图像绝大部分信息，而且指标的计算只需要得到原始数据的协方差阵，而不必对原始数据进行真正的主成分变换，极大的降低了计算量。贝叶斯和K-均值分类实验表明．该算法是有效可行的。     

Radiometric normalization, compositing, and quality control forsatellite high resolution image mosaics over large areas

An objective normalization procedure has been developed to create image mosaics of radiometric equalization radiometric normalization for image mosaics (RNIM). The procedure employs a band-specific principal component analysis for overlap areas to achieve accurate and consistent radiometric transforms in each spectral band. It is demonstrated that the result of radiometric equalization is independent of the order of images to be mosaicked after the radiometric normalization adjustment is made. The selection of corresponding pixel pairs in the overlap area is controlled by using band-specific linear correlation coefficients, and the criteria for rejecting the cloudy and land-cover changed pixels. The final result is controlled quantitatively by employing the first and second principal components for the input data, which in turn depend on the selection of corresponding pixel pairs in the overlap area. In general, the radiometric resolution of input images can be conserved as long as gain ⩾1 and offset ⩾0 because of the stored format of the unsigned integer. The RNIM procedure accommodates these conditions. To take the best advantage of the data in the overlap areas, a pixel-based composite technique is employed in the production of the final mosaic. The selection of corresponding pixel pairs and the final result can be controlled and assessed with quantitative criteria. Therefore, this approach produces an objective, analyst-independent result and can be automated. The method has been successfully applied to six Landsat TM images of the BOREAS transect in Saskatchewan and Manitoba, Canada     

Hyperspectral Fluorescence Imaging for Mouse Skin Tumor Detection

This paper presents a hyperspectral imaging technique based on laser‐induced fluorescence for non‐invasive detection of tumorous tissue on mouse skin. Hyperspectral imaging sensors collect image data in a number of narrow, adjacent spectral bands. Such high‐resolution measurement of spectral information reveals contiguous emission spectra at each image pixel useful for the characterization of constituent materials. The hyperspectral image data used in this study are fluorescence images of mouse skin consisting of 21 spectral bands in the visible spectrum of the wavelengths ranging from 440 nm to 640 nm. Fluorescence signal is measured with the use of laser excitation at 337 nm. An acousto‐optic tunable filter (AOTF) is used to capture images at 10 nm intervals. All spectral band images are spatially registered with the reference band image at 490 nm to obtain exact pixel correspondences by compensating the spatial offsets caused by the refraction differences in AOTF at different wavelengths during the image capture procedure. The unique fluorescence spectral signatures demonstrate a good separation to differentiate malignant tumors from normal tissues for rapid detection of skin cancers without biopsy.     

p值统计量建模独立性的高光谱波段选择方法

近年来，p值统计量的使用规范引起了统计学界的极大关注和集中讨论，广泛认为，p值统计量可表达观测数据与备择假设之间的不相容程度。为探究高光谱图像波段的相关分析p值与其样本独立性的联系，进行了演绎推理和实例验证，研究表明，与相关系数r统计量相比，相关分析p统计量可直接表达波段样本的独立性，且p值矩阵具有高水平的自稀疏性，便于建模和计算。进而，对相关性p值矩阵进行直方图频数统计，提出一种基于p值的高光谱自适应波段选择方法pSMBS。选取典型数据进行了监督分类实验，结果表明，在Kappa系数、总体精度（OA）和平均精度（AA）上，pSMBS均优于同类方法ABS、InfFS和LSFS。说明pSMBS在高光谱波段选择方面具有突出的有效性，这也佐证了相关性p值对波段独立性的强表征能力。     

Super-resolution reconstruction of hyperspectral images.

Hyperspectral images are used for aerial and space imagery applications, including target detection, tracking, agricultural, and natural resource exploration. Unfortunately, atmospheric scattering, secondary illumination, changing viewing angles, and sensor noise degrade the quality of these images. Improving their resolution has a high payoff, but applying super-resolution techniques separately to every spectral band is problematic for two main reasons. First, the number of spectral bands can be in the hundreds, which increases the computational load excessively. Second, considering the bands separately does not make use of the information that is present across them. Furthermore, separate band super-resolution does not make use of the inherent low dimensionality of the spectral data, which can effectively be used to improve the robustness against noise. In this paper, we introduce a novel super-resolution method for hyperspectral images. An integral part of our work is to model the hyperspectral image acquisition process. We propose a model that enables us to represent the hyperspectral observations from different wavelengths as weighted linear combinations of a small number of basis image planes. Then, a method for applying super resolution to hyperspectral images using this model is presented. The method fuses information from multiple observations and spectral bands to improve spatial resolution and reconstruct the spectrum of the observed scene as a combination of a small number of spectral basis functions.     

An information-theoretic approach to spectral variability,similarity, and discrimination for hyperspectral image analysis

A hyperspectral image can be considered as an image cube where the third dimension is the spectral domain represented by hundreds of spectral wavelengths. As a result, a hyperspectral image pixel is actually a column vector with dimension equal to the number of spectral bands and contains valuable spectral information that can be used to account for pixel variability, similarity and discrimination. We present a new hyperspectral measure, the spectral information measure (SIM), to describe spectral variability and two criteria, spectral information divergence and spectral discriminatory probability for spectral similarity and discrimination, respectively. The spectral information measure is an information-theoretic measure which treats each pixel as a random variable using its spectral signature histogram as the desired probability distribution. Spectral information divergence (SID) compares the similarity between two pixels by measuring the probabilistic discrepancy between two corresponding spectral signatures. The spectral discriminatory probability calculates spectral probabilities of a spectral database (library) relative to a pixel to be identified so as to achieve material identification. In order to compare the discriminatory power of one spectral measure relative to another, a criterion is also introduced for performance evaluation, which is based on the power of discriminating one pixel from another relative to a reference pixel. The experimental results demonstrate that the new hyperspectral measure can characterize spectral variability more effectively than the commonly used spectral angle mapper (SAM)