Deep learning(DL)has shown its superior performance in dealing with various computer vision tasks in recent years.As a simple and effective DL model,autoencoder(AE)is popularly used to decompose hyperspectral images(H...Deep learning(DL)has shown its superior performance in dealing with various computer vision tasks in recent years.As a simple and effective DL model,autoencoder(AE)is popularly used to decompose hyperspectral images(HSIs)due to its powerful ability of feature extraction and data reconstruction.However,most existing AE-based unmixing algorithms usually ignore the spatial information of HSIs.To solve this problem,a hypergraph regularized deep autoencoder(HGAE)is proposed for unmixing.Firstly,the traditional AE architecture is specifically improved as an unsupervised unmixing framework.Secondly,hypergraph learning is employed to reformulate the loss function,which facilitates the expression of high-order similarity among locally neighboring pixels and promotes the consistency of their abundances.Moreover,L_(1/2)norm is further used to enhance abundances sparsity.Finally,the experiments on simulated data,real hyperspectral remote sensing images,and textile cloth images are used to verify that the proposed method can perform better than several state-of-the-art unmixing algorithms.展开更多
由于不同的照明条件、复杂的大气环境等因素,相同端元的光谱特征在图像的不同位置呈现出可见的差异,这种现象被称为端元的光谱变异性。在相当大的场景中,端元的变异性可能很大,但在适度的局部同质区内,变异性往往很小。扰动线性混合模型...由于不同的照明条件、复杂的大气环境等因素,相同端元的光谱特征在图像的不同位置呈现出可见的差异,这种现象被称为端元的光谱变异性。在相当大的场景中,端元的变异性可能很大,但在适度的局部同质区内,变异性往往很小。扰动线性混合模型(Perturbed Linear Mixing Model,PLMM)在解混的过程中可以减轻端元变异性造成的不利影响,但是对缩放效应造成的变异性的处理能力较弱。为此,本文改进了扰动线性混合模型,引入了尺度因子以处理缩放效应造成的变异性,并结合超像素分割算法划分局部同质区,然后设计出基于局部同质区共享端元变异性的解混算法(Shared Endmember Variability in Unmixing,SEVU)。与扰动线性混合模型,扩展线性混合模型(Extended Linear Mixing Model,ELMM)等算法相比,所提SEVU算法在合成数据集上平均端元光谱角距离(mean Spectral Angle Distance,mSAD)和丰度均方根误差(abundance Root Mean Square Error,aRMSE)最优,分别为0.0855和0.0562;在Jasper Ridge和Cuprite真实数据集上mSAD是最优的,分别为0.0603和0.1003。在合成数据集和两个实测数据集上的实验结果验证了SEVU算法的有效性。展开更多
Chaetoceros Ehrenberg is one of the most diverse genera of planktonic diatoms.The species in section Chaetoceros are characterized by cells and setae having numerous chloroplasts and being widely distributed.However,t...Chaetoceros Ehrenberg is one of the most diverse genera of planktonic diatoms.The species in section Chaetoceros are characterized by cells and setae having numerous chloroplasts and being widely distributed.However,the delimitations of some species are problematic because of limited morphological information in the classical descriptions.Monoclonal strains of the section Chaetoceros were established,morphological features were studied using light and electron microscopy,and the hypervariable D 1-D 3 region of the nuclear ribosomal large subunit gene was sequenced to address phylogenetic relationships.Fifteen species belonging to the section Chaetoceros were recorded,including two new species,C.hainanensis sp.nov.and C.tridiscus sp.nov.Chaetoceros hainanensis was characterized by straight chains,narrowly lanceolate to hexagonal apertures,sibling setae diverging in nearly right angles,stipule-shaped spines on terminal setae and arrowhead-shaped spines on intercalary setae.C.tridiscus had short straight chains,narrowly lanceolate apertures,arrowhead-shaped spines and circular poroids arranged in a grid pattern on terminal and intercalary setae.The phylogenetic analyses revealed six groups formed by 19 species within the section Chaetoceros,which was found to be monophyletic.The subdivision of the section is still not well understood.The morphological characters within each group varied considerably and molecular information on more species are needed to enrich the phylogenetic profiling.展开更多
基金National Natural Science Foundation of China(No.62001098)Fundamental Research Funds for the Central Universities of Ministry of Education of China(No.2232020D-33)。
文摘Deep learning(DL)has shown its superior performance in dealing with various computer vision tasks in recent years.As a simple and effective DL model,autoencoder(AE)is popularly used to decompose hyperspectral images(HSIs)due to its powerful ability of feature extraction and data reconstruction.However,most existing AE-based unmixing algorithms usually ignore the spatial information of HSIs.To solve this problem,a hypergraph regularized deep autoencoder(HGAE)is proposed for unmixing.Firstly,the traditional AE architecture is specifically improved as an unsupervised unmixing framework.Secondly,hypergraph learning is employed to reformulate the loss function,which facilitates the expression of high-order similarity among locally neighboring pixels and promotes the consistency of their abundances.Moreover,L_(1/2)norm is further used to enhance abundances sparsity.Finally,the experiments on simulated data,real hyperspectral remote sensing images,and textile cloth images are used to verify that the proposed method can perform better than several state-of-the-art unmixing algorithms.
文摘由于不同的照明条件、复杂的大气环境等因素,相同端元的光谱特征在图像的不同位置呈现出可见的差异,这种现象被称为端元的光谱变异性。在相当大的场景中,端元的变异性可能很大,但在适度的局部同质区内,变异性往往很小。扰动线性混合模型(Perturbed Linear Mixing Model,PLMM)在解混的过程中可以减轻端元变异性造成的不利影响,但是对缩放效应造成的变异性的处理能力较弱。为此,本文改进了扰动线性混合模型,引入了尺度因子以处理缩放效应造成的变异性,并结合超像素分割算法划分局部同质区,然后设计出基于局部同质区共享端元变异性的解混算法(Shared Endmember Variability in Unmixing,SEVU)。与扰动线性混合模型,扩展线性混合模型(Extended Linear Mixing Model,ELMM)等算法相比,所提SEVU算法在合成数据集上平均端元光谱角距离(mean Spectral Angle Distance,mSAD)和丰度均方根误差(abundance Root Mean Square Error,aRMSE)最优,分别为0.0855和0.0562;在Jasper Ridge和Cuprite真实数据集上mSAD是最优的,分别为0.0603和0.1003。在合成数据集和两个实测数据集上的实验结果验证了SEVU算法的有效性。
基金Supported by the Joint Fund of National Natural Science Foundation of China and Chinese Shandong Province(No.U 2106205)the National Natural Science Foundation of China(No.32170206)the National Key Research and Development Program of China(No.2022YFC3105201)。
文摘Chaetoceros Ehrenberg is one of the most diverse genera of planktonic diatoms.The species in section Chaetoceros are characterized by cells and setae having numerous chloroplasts and being widely distributed.However,the delimitations of some species are problematic because of limited morphological information in the classical descriptions.Monoclonal strains of the section Chaetoceros were established,morphological features were studied using light and electron microscopy,and the hypervariable D 1-D 3 region of the nuclear ribosomal large subunit gene was sequenced to address phylogenetic relationships.Fifteen species belonging to the section Chaetoceros were recorded,including two new species,C.hainanensis sp.nov.and C.tridiscus sp.nov.Chaetoceros hainanensis was characterized by straight chains,narrowly lanceolate to hexagonal apertures,sibling setae diverging in nearly right angles,stipule-shaped spines on terminal setae and arrowhead-shaped spines on intercalary setae.C.tridiscus had short straight chains,narrowly lanceolate apertures,arrowhead-shaped spines and circular poroids arranged in a grid pattern on terminal and intercalary setae.The phylogenetic analyses revealed six groups formed by 19 species within the section Chaetoceros,which was found to be monophyletic.The subdivision of the section is still not well understood.The morphological characters within each group varied considerably and molecular information on more species are needed to enrich the phylogenetic profiling.
