高光谱图像具有光谱分辨率高与“ 图谱合一”的特点，被广泛应用于环境监测、精 准农业以及地质勘探等领域。然而，由于仪器空间分辨率低及复杂的成像条件等限制， 高光谱数据中存在大量的混合像元，为地物信息的解译带来了重重阻碍。因此，研究高 光谱图像混合像元分解“（解混）题题具有非常重要的现实意义。解混是处理混合像元题 题最有效的分析方法，旨在从高光谱数据中分离出混合物质的光谱（端元）及其相应比 例（丰度）。近年来，深度学习的迅速发展极大推动了解混技术。虽然基于深度学习的 解混方法在信息挖掘和泛化性能方面相较于传统方法具有明显优势，但是目前基于深度 学习的解混方法主要关注于图像光谱维信息，而空间维信息的利用方面仍然停留在使用 堆叠卷积、滤波等网络层的表层处理阶段，容易造成信息损失等题题，限制其解混精度。 鉴于此，本研究基于深度学习理论，引入 Transformer 网络构建深度解混网络，研究高 光谱图像空-谱信息对解混性能的影响，并进一步从实际应用角度提高解混模型的泛化 性。主要工作内容如下： (1) 提出了一种基于深度嵌入式 Transformer 网络的空-谱解混方法。针对当前高光 谱解混研究中图像空-谱信息获取不充分的题题，该方法以自动编码器网络为基本构架， 受到 Transformer in Transformer（TNT）模型的启发，使用内外嵌入式 Transformer 网络 构建深度解混网络，实现了局部与整体空-谱信息共享，为解混提供了图像深层空-谱信 息。同时，该方法涉及少量卷积运算，避免了网络训练时可能出现的信息损失与过拟合 现象，提高了解混精度。最后，采用模拟数据集和经典真实高光谱数据集进行解混实验， 同时，使用常用的精度评估指标评定该方法的解混精度，并与其他经典解混方法对比， 结果显示该方法的解混精度皆有所提高。最后，使用真实项目中的高光谱数据集进行解 混应用实验，通过定性分析该方法的解混效果较符合真实地物的分布情况，具有一定的 实际应用价值。实验结果表明，基于深度嵌入式 Transformer 网络的空-谱解混方法能够 充分利用图像的空-谱信息提升网络的解混性能。 (2) 提出了一种自适应分块的深度嵌入式 Transformer 网络的空-谱解混方法。该方 法同样从空-谱协同角度出发，针对 Transformer 网络无法对长宽不一致的高光谱图像进 行解混的题题，通过 resize 操作使得不均匀分割的图块能够自适应生成相同长度的图块 序列，并输入内外嵌入式 Transformer 结构中获取图像深层空-谱信息，从而实现对长宽 不一致的自然高光谱图像解混的目的，提高模型的泛化性。同时，该方法使用 CNN 与 Transformer 网络结合的策略，减少冗余波段，改善训练参数多、计算复杂的题题。最后， 同样采用模拟数据集和经典真实高光谱数据集进行解混实验，评估该方法的解混精度并 与同类经典解混方法对比，结果显示该方法皆能获得相对较高的解混精度。最后，使用 真实项目中长宽尺寸不一致的高光谱图像进行解混实验，通过定性分析该方法具有一定 的实际应用价值且普适性更高。实验结果表明，自适应分块的深度嵌入式 Transformer 网 络的空-谱解混方法能够利用图像的空-谱信息提高解混精度，且该方法应用于复杂场景 下的解混效果更佳。

Hyperspectral imagery, characterized by high spectral resolution and the integration of spatial and spectral information, is widely applied in fields such as environmental monitoring, precision agriculture, and geological exploration. However, due to limitations in instrument spatial resolution and complex imaging conditions, hyperspectral data often contain numerous mixed pixels, posing significant challenges to the interpretation of terrestrial information. Thus, research on hyperspectral image mixed pixel decomposition (unmixing) is of great practical significance. Unmixing is the most effective analytical method for addressing the mixed pixel problem, aiming to separate the spectra of mixed substances (endmembers) and their corresponding proportions (abundances) from hyperspectral data. In recent years, the rapid development of deep learning has significantly advanced unmixing techniques. Although deep learning-based unmixing methods have demonstrated clear advantages in information extraction and generalization performance compared to traditional methods, current approaches primarily focus on spectral information in images. The utilization of spatial information remains at a superficial level, such as using stacked convolutional layers and filters, which can lead to information loss and limit unmixing accuracy. In light of this, our study introduces a deep unmixing network based on deep learning theory and the Transformer network, investigating the impact of spatial-spectral information on unmixing performance and further enhancing the generalization of the unmixing model from a practical application perspective. The main contributions of this research are as follows: (1) Aspatial-spectral unmixing method based on a Deep Embedded Transformer Network. Addressing the insufficient acquisition of spatial-spectral information in current hyperspectral unmixing research, this method uses an autoencoder network as the basic framework, inspired by the Transformer in Transformer (TNT) model. It employs nested embedded Transformer networks to construct a deep unmixing network, achieving local and global spatial-spectral information sharing and providing deep spatial-spectral information for unmixing. This approach involves minimal convolution operations, avoiding potential information loss and overfitting during network training, thus improving unmixing accuracy. Finally, unmixing experiments are conducted using simulated datasets and classic real hyperspectral datasets. The accuracy of this method is evaluated using common precision assessment metrics and compared with other classical unmixing methods, showing improved unmixing accuracy. Lastly, a practical unmixing application experiment using a real-project hyperspectral dataset qualitatively demonstrated that the unmixing results aligned well with the actual distribution of terrestrial features, indicating practical application value. Experimental results show that the spatial-spectral unmixing method based on deep embedded Transformer networks effectively utilizes spatial-spectral information to enhance network unmixing performance. (2) Adaptable Block Deep Embedded Transformer Network for spatial-spectral unmixing. This method also approaches from a spatial-spectral synergy perspective, addressing the issue of Transformer networks being unable to unmix hyperspectral images with inconsistent width and height. Through resizing operations, unevenly segmented image blocks are adaptively generated into block sequences of the same length and input into the nested embedded Transformer structure to obtain deep spatial-spectral information, achieving unmixing of natural hyperspectral images with inconsistent dimensions and improving model generalization. Additionally, this method combines CNN and Transformer networks to reduce redundant bands and alleviate the problem of excessive training parameters and computational complexity. Finally, unmixing experiments are conducted using simulated datasets and classic real hyperspectral datasets, evaluating the unmixing accuracy of this method and comparing it with similar classical unmixing methods, showing relatively high unmixing accuracy. Lastly, a practical unmixing experiment using hyperspectral images with inconsistent dimensions from real projects qualitatively demonstrated that this method has higher practical application value and generality. Experimental results show that the adaptable block deep embedded Transformer network for spatial-spectral unmixing effectively utilizes spatial-spectral information to improve unmixing accuracy, and performs better in complex scenarios.

[1] 张兵. 高光谱图像处理与信息提取前沿[J]. 遥感学报, 2016, 20(5):1062-1090.

[2] 苏红军. 高光谱遥感图像降维: 进展, 挑战与展望[J]. 遥感学报, 2022, 26(8): 1504-1529.

[3] 张淳民, 穆廷魁, 颜廷昱, 等. 高光谱遥感技术发展与展望[J]. 航天返回与遥感, 2018, 39(3): 104-14.

[4] 龙珍, 刘翼鹏, 苟艺馨, 等. 高光谱图像的线性张量子空间模型及降噪应用[J]. 中国图象图形学报, 2023, 28(8): 2505-2521.

[5] Su Y, Jiang M, Gao L, et al. Graph-Cut-Based Collaborative Node Embeddings for Hyperspectral Images Classification[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19: 1-5.

[6] Zhao Z, Xu X, Li S, et al. Hyperspectral Image Classification Using Groupwise Separable Convolutional Vision Transformer Network[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 1-17.

[7] Su Y, Gao L, Jiang M, et al. NSCKL: Normalized Spectral Clustering With Kernel-Based Learning for Semisupervised Hyperspectral Image Classification[J]. IEEE Transactions on Cybernetics, 2023, 53(10): 6649-6662.

[8] 涂坤, 熊凤超, 陆建峰, 等. 多任务的高光谱图像卷积稀疏编码去噪网络[J]. 中国图象图形学报, 2024, 29(1): 280-292.

[9] 叶珍, 白璘, 何明一. 高光谱图像空谱特征提取综述[J]. 中国图象图形学报, 2021, 26(8): 1737-1763.

[10] Fang T, Zhu F, Chen J. Hyperspectral Unmixing Based on Multilinear Mixing Model Using Convolutional Autoencoders[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 1-16.

[11] Zhao M, Chen J, Dobigeon N. AE-RED: A Hyperspectral Unmixing Framework Powered by Deep Autoencoder and Regularization by Denoising[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 1-15.

[12] 蓝金辉, 邹金霖, 郝彦爽, 等. 高光谱遥感图像混合像元分解研究进展[J]. 遥感学报, 2018, 22(01): 13-27.

[13] 靳淇文, 马泳, 樊凡, 等. 生成对抗网络的无监督高光谱解混[J]. 遥感学报, 2023, 27(08): 1964-1974.

[14] Zhao M, Shi S, Chen J, et al. A 3-D-CNN framework for hyperspectral unmixing with spectral variability[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 1-14.

[15] 冯如意,王力哲,曾铁勇.高光谱遥感图像亚像元信息提取方法综述[J]. 测绘学报, 2023, 52(07): 1187-1201.

[16] 李思佳. 基于深度学习的空谱联合高光谱解混方法研究[D]. 郑州: 郑州轻工业大学, 2024.

[17] Su Y, Marinoni A, Li J, et al. Stacked nonnegative sparse autoencoders for robust hyperspectral unmixing[J]. IEEE Geoscience and Remote Sensing Letters, 2018, 15(9): 1427-1431.

[18] 朱治青, 苏远超, 李朋飞, 等. 双约束深度卷积网络的高光谱图像空谱解混方法[J]. 信号处理, 2023, 39(1): 128-142.

[19] 张作宇, 廖守亿, 孙大为, 等. 稀疏差异先验信息支持的高光谱图像稀疏解混算法[J].测绘学报, 2020, 49(08): 1032-1041.

[20] 刘帅, 邢光龙. 高光谱图像混合像元多维卷积网络协同分解法[J]. 测绘学报, 2020, 49(12): 1600-1608.

[21] You X, Su Y, Gao L, et al. DSET-Net: deep self-embedded transformer network for hyperspectral unmixing[C]// DSET-Net: deep self-embedded transformer network for hyperspectral unmixing. Fourth International Conference on Geology, Mapping, and Remote Sensing (ICGMRS 2023). 12978: 110-115.

[22] 苏远超, 许若晴, 高连如, 等. 基于深度学习的高光谱遥感图像混合像元分解研究综述[J]. 遥感学报, 2024, 28(1): 1-19.

[23] 王毓乾.基于空间-光谱分析的高光谱遥感影像稀疏解混研究[J]. 测绘学报, 2017, 46(08): 1072.

[24] Su Y, Zhu Z, Gao L, et al. DAAN: A Deep Autoencoder-Based Augmented Network for Blind Multilinear Hyperspectral Unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 1-15.

[25] 雷明雨. 基于深度学习和空谱联合先验的高光谱遥感解混研究[D]. 西安: 西安电子科技大学, 2022.

[26] Su Y, Xu X, Li J, et al. Deep autoencoders with multitask learning for bilinear hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 59(10): 8615-8629.

[27] Zhao M, Yan L, Chen J. LSTM-DNN based autoencoder network for nonlinear hyperspectral image unmixing[J]. IEEE Journal of Selected Topics in Signal Processing, 2021, 15(2): 295-309.

[28] 彭倩, 张兵, 孙旭, 等. 结合空间光谱预处理和约束非负矩阵分解的高光谱图像混合像元分解[J]. 中国图像图形学, 2017, 22(04): 542-550.

[29] Bioucas-Dias J M, Plaza A, Dobigeon N, et al. Hyperspectral unmixing overview: Geometrical, statistical, and sparse regression-based approaches[J]. IEEE journal of selected topics in applied earth observations and remote sensing, 2012, 5(2): 354-379.

[30] Palsson B, Sigurdsson J, Sveinsson J R, et al. Hyperspectral unmixing using a neural network autoencoder[J]. IEEE Access, 2018, 6: 25646-25656.

[31] Jin Q, Ma Y, Fan F, et al. Adversarial autoencoder network for hyperspectral unmixing[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 34(8): 4555-4569.

[32] 宋晓瑞, 吴玲达, 孟祥利. 利用稳健非负矩阵分解实现无监督高光谱解混[J]. 中国图像图形学报, 2020, 25(04): 801-812.

[33] 张兵. 高光谱图像混合像元分解[M]. 科学出版社, 2015.

[34] 黄勇发. 基于自编码网络的高光谱解混方法研究[D]. 西安: 西安电子科技大学, 2022.

[35] 李二森, 朱述龙, 周晓明, 等. 高光谱图像端元提取方法研究进展与比较[J]. 遥感学报, 2011, 15(4): 659-679.

[36] Boardman J W, Kruse F A, Green R O. Mapping target signatures via partial unmixing of AVIRIS data[C]//Summaries of the fifth annual JPL airborne earth science workshop. Volume 1: AVIRIS workshop. 1995.

[37] Nascimento J M P, Dias J M B. Vertex component analysis: A fast method to unmix hyperspectral data[J]. IEEE transactions on Geoscience and Remote Sensing, 2005, 43(4): 898-910.

[38] Winter M E. N-FINDR: An algorithm for fast autonomous spectral end-member determination in hyperspectral data[C]//Imaging spectrometry V. SPIE, 1999, 3753: 266-275.

[39] Li J, Bioucas-Dias J M. Minimum volume simplex analysis: A fast algorithm to unmix hyperspectral data[C]//IGARSS 2008-2008 IEEE International Geoscience and Remote Sensing Symposium. IEEE, 2008, 3: 250-253.

[40] Chan T H, Chi C Y, Huang Y M, et al. A convex analysis-based minimum-volume enclosing simplex algorithm for hyperspectral unmixing[J]. IEEE Transactions on Signal Processing, 2009, 57(11): 4418-4432.

[41] Bioucas-Dias J M. A variable splitting augmented Lagrangian approach to linear spectral unmixing[C]//2009 First workshop on hyperspectral image and signal processing: Evolution in remote sensing. IEEE, 2009: 1-4.

[42] 谢金凤, 陈涛. 考虑光谱信息和超像素分割的高光谱解混网络[J]. 遥感学报, 2024, 28(01): 142-153.

[43] Stone J V. Independent component analysis: a tutorial introduction[M]. Massachusetts Institute of Technology, 2004.

[44] Liu W, Mei X, Ma Y, et al. Graph l₁-Laplacians Regularized GMM for Hyperspectral Unmixing[J]. IEEE Geoscience and Remote Sensing Letters, 2023, 20: 1-5.

[45] Liu H, Lu Y, Wu Z, et al. Spectral Variability Bayesian Unmixing for Hyperspectral Sequence in Wavelet Domain[J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61: 1-19.

[46] Feng X R, Li H C, Wang R, et al. Hyperspectral unmixing based on nonnegative matrix factorization: A comprehensive review[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, 15: 4414-4436.

[47] Miao L, Qi H. Endmember extraction from highly mixed data using minimum volume constrained nonnegative matrix factorization[J]. IEEE Transactions on Geoscience and Remote Sensing, 2007, 45(3): 765-777.

[48] Qian Y, Jia S, Zhou J, et al. Hyperspectral unmixing via L1/2 sparsity-constrained nonnegative matrix factorization[J]. IEEE Transactions on Geoscience and Remote Sensing, 2011, 49(11): 4282-4297.

[49] Li J, Bioucas-Dias J M, Plaza A, et al. Robust collaborative nonnegative matrix factorization for hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(10): 6076-6090.

[50] Qian Y, Xiong F, Zeng S, et al. Matrix-vector nonnegative tensor factorization for blind unmixing of hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 55(3): 1776-1792.

[51] Wang X, Zhong Y, Zhang L, et al. Spatial group sparsity regularized nonnegative matrix factorization for hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2017, 55(11): 6287-6304.

[52] Dong L, Yuan Y, Luxs X. Spectral-spatial joint sparse NMF for hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 59(3): 2391-2402.

[53] Shu S, Huang T-Z, Huang J, et al. Endmembers compression based nonnegative matrix factorization for hyperspectral unmixing[J]. International Journal of Remote Sensing, 2023, 44(6): 2064-2092.

[54] Ayres L C, Borsoi R A, Bermudez J C M, et al. A Generalized Multiscale Bundle-Based Hyperspectral Sparse Unmixing Method[J]. IEEE Geoscience and Remote Sensing Letters, 2024, 21: 1-5.

[55] Bioucas-Dias J M, Figueiredo M A T. Alternating direction algorithm for constrained sparse regression: Application to hyperspectral unmixing[C]//2010 2nd Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing. IEEE, 2010: 1-4.

[56] 汪顺清, 杨劲翔, 邵远天, 等. 高光谱图像非负稀疏分量分解建模与鲁棒性解混方法[J]. 中国图象图形学报, 2023, 28(02): 613-627.

[57] Iordache M-D, Bioucas-Dias J M, Plaza A. Sparse unmixing of hyperspectral data[J]. IEEE Transactions on Geoscience and Remote Sensing, 2011, 49(6): 2014-2039.

[58] Kumar U, Milesi C, Nemani R R, et al. Sparse unmixing via variable splitting and augmented Lagrangian for vegetation and urban area classification using Landsat data[J]. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2015, 40: 59-65.

[59] Li H, Feng R, Wang L, et al. Superpixel-based reweighted low-rank and total variation sparse unmixing for hyperspectral remote sensing imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 59(1): 629-647.

[60] Xu X, Pan B, Chen Z, et al. Simultaneously multiobjective sparse unmixing and library pruning for hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 59(4): 3383-3395.

[61] Chen T, Liu Y, Zhang Y, et al. Superpixel-based collaborative and low-rank regularization for sparse hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 1-16.

[62] Guo R, Wang W, Qi H. Hyperspectral image unmixing using autoencoder cascade[C]//2015 7th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS). IEEE, 2015: 1-4.

[63] Su Y, Li J, Plaza A, et al. DAEN: Deep autoencoder networks for hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(7): 4309-4321.

[64] Qu Y, Qi H. uDAS: An untied denoising autoencoder with sparsity for spectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 57(3): 1698-1712.

[65] Ozkan S, Kaya B, Akar G B. Endnet: Sparse autoencoder network for endmember extraction and hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 57(1): 482-496.

[66] Borsoi R A, Imbiriba T, Bermudez J C M. Deep generative endmember modeling: An application to unsupervised spectral unmixing[J]. IEEE Transactions on Computational Imaging, 2019, 6: 374-384.

[67] Jin Q, Ma Y, Mei X, et al. TANet: An unsupervised two-stream autoencoder network for hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 60: 1-15.

[68] Palsson B, Ulfarsson M O, Sveinsson J R. Convolutional autoencoder for spectral-spatial hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59(1): 535-549.

[69] Gao L, Han Z, Hong D, et al. CyCU-Net: Cycle-consistency unmixing network by learning cascaded autoencoders[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 60: 1-14.

[70] Fang Y, Wang Y, Xu L, et al. BCUN: Bayesian fully convolutional neural network for hyperspectral spectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 1-14.

[71] Qi L, Chen Z, Gao F, et al. Multiview spatial-spectral two-stream network for hyperspectral image unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61: 1-16.

[72] Wang B, Yao H, Song D, et al. SSF-Net: A Spatial-Spectral Features Integrated Autoencoder Network for Hyperspectral Unmixing[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024, 17: 1781-1794.

[73] Zhu Z, Su Y, Jiang M, et al. Spectral-Spatial Hyperspectral Unmixing Using Double-Constraints Convolutional Autoencoder[C]//2023 13th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS). IEEE, 2023: 1-4.

[74] Hua Z, Li X, Jiang J, et al. Gated autoencoder network for spectral-spatial hyperspectral unmixing[J]. Remote Sensing, 2021, 13(16): 3147.

[75] Ghosh P, Roy S K, Koirala B, et al. Hyperspectral unmixing using transformer network[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 1-16.

[76] Han K, Xiao A, Wu E, et al. Transformer in transformer[J]. Advances in neural information processing systems, 2021, 34: 15908-15919.

[77] Iordache M D, Bioucas-Dias J M, Plaza A. Collaborative sparse regression for hyperspectral unmixing[J]. IEEE Transactions on geoscience and remote sensing, 2013, 52(1): 341-354.

[78] Lemme A, Reinhart R F, Steil J J. Online learning and generalization of parts-based image representations by non-negative sparse autoencoders[J]. Neural Networks, 2012, 33: 194-203.

[79] Vargas-Guzmán J A, Dimitrakopoulos R. Computational properties of min/max autocorrelation factors[J]. Computers & Geosciences, 2003, 29(6): 715-723.

[80] Calhoun V D, Adali T. Unmixing fMRI with independent component analysis[J]. IEEE Engineering in Medicine and Biology Magazine, 2006, 25(2): 79-90.

[81] Heinz D C. Fully constrained least squares linear spectral mixture analysis method for material quantification in hyperspectral imagery[J]. IEEE transactions on geoscience and remote sensing, 2001, 39(3): 529-545.

[82] Kim K I, Franz M O, Scholkopf B. Iterative kernel principal component analysis for image modeling[J]. IEEE transactions on pattern analysis and machine intelligence, 2005, 27(9): 1351-1366.

[83] Heylen R, Burazerovic D, Scheunders P. Fully constrained least squares spectral unmixing by simplex projection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2011, 49(11): 4112-4122.

[84] Chen J, Zhao M, Wang X, et al. Integration of physics-based and data-driven models for hyperspectral image unmixing: A summary of current methods[J]. IEEE Signal Processing Magazine, 2023, 40(2): 61-74.

[85] Hua Z, Li X, Qiu Q, et al. Autoencoder network for hyperspectral unmixing with adaptive abundance smoothing[J]. IEEE Geoscience and Remote Sensing Letters, 2020, 18(9): 1640-1644.

[86] 苏远超. 基于深度自动编码器的高光谱遥感图像混合像元分解研究[D]. 广州: 中山大学, 2019.

[87] Zhang X, Sun Y, Zhang J, et al. Hyperspectral unmixing via deep convolutional neural networks[J]. IEEE Geoscience and Remote Sensing Letters, 2018, 15(11): 1755-1759.

[88] Jin B, Zhu Y, Huang W, et al. An Efficient Attention-Based Convolutional Neural Network That Reduces the Effects of Spectral Variability for Hyperspectral Unmixing[J]. Applied Sciences, 2022, 12(23): 1-20.

[89] Chua L O. CNN: A vision of complexity[J]. International Journal of Bifurcation and Chaos, 1997, 7(10): 2219-2425.

[90] Hadi F, Yang J, Farooque G, et al. Deep convolutional transformer network for hyperspectral unmixing[J]. European Journal of Remote Sensing, 2023, 56(1): 1-18.

[91] Zhou W, Kamata S, Wang H, et al. Segmented Recurrent Transformer with Cubed 3D-Multiscanning Strategy for Hyperspectral Image Classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 1-22.

[92] Wang Y, Shi S, Chen J. Efficient Blind Hyperspectral Unmixing with Non-Local Spatial Information Based on Swin Transformer[C]//IGARSS 2023-2023 IEEE International Geoscience and Remote Sensing Symposium. IEEE, 2023: 5898-5901.

[93] Han Z, Hong D, Gao L, et al. Multimodal hyperspectral unmixing: Insights from attention networks[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 1-13.

[94] Jiang M, Liu G, Su Y, et al. Self-attention empowered graph convolutional network for structure learning and node embedding[J]. Pattern Recognition, 2024, 153: 1-11.

[95] He X, Zhou Y, Zhao J, et al. Swin transformer embedding UNet for remote sensing image semantic segmentation[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 1-15.

[96] Wang L, Zhang X, Zhang J, et al. Pixel-to-Abundance Translation: Conditional Generative Adversarial Networks Based on Patch Transformer for Hyperspectral Unmixing[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024, 17: 5734-5749.

[97] Duan Y, Xu X, Li T, et al. UnDAT: Double-Aware Transformer for Hyperspectral Unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61: 1-12.

[98] 李妙宇, 付莹. 用于多光谱和高光谱图像融合的联合自注意力Transformer[J]. 中国图象图形学报, 2023, 28(12): 3922-3934.

[99] Bhakthan S M, Loganathan A. A hyperspectral unmixing model using convolutional vision transformer[J]. Earth Science Informatics, 2024: 1-19.

[100] Yang Z, Xu M, Liu S, et al. UST-Net: A U-Shaped Transformer Network Using Shifted Windows for Hyperspectral Unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61: 1-15.

[101]Chen J, Yang C, Zhang L, et al. TCCU-Net: Transformer and CNN Collaborative Unmixing Network for Hyperspectral image[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024, 17: 8073-8099.

[102]Rasti B, Koirala B, Scheunders P, et al. UnDIP: Hyperspectral unmixing using deep image prior[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 60: 1-15.

[103]童庆禧, 张兵, 张立福. 中国高光谱遥感的前沿进展[J]. 遥感学报, 2016, 20(5): 689-707.

[104]Beyer L, Izmailov P, Kolesnikov A, et al. Flexivit: One model for all patch sizes[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023: 14496-14506.

[105]Sun B, Su Y, Sun H, et al. Generative Adversarial Autoencoder Network for Anti-Shadow Hyperspectral Unmixing[J]. IEEE Geoscience and Remote Sensing Letters, 2024, 21: 1-5.