基于组合核函数相关向量机的高光谱图像分类
Hyperspectral Image Classification Based on Composite Kernel Relevance Vector Machine
DOI: 10.12677/CSA.2023.137138, PDF,    科研立项经费支持
作者: 孙 琤*：中国五洲工程设计集团有限公司，北京；刘东好：中兵勘察设计研究院有限公司，北京
关键词: 图像分类；相关向量机；组合核函数；多尺度形态学特征；高光谱图像；Image Classification； Relevance Vector Machine； Composite Kernel； Multi-Scale Morphological Pro?les； Hyperspectral Image
摘要: 本文提出了一种基于组合核函数相关向量机的高光谱图像分类方法。基于核函数的特性构造出三种形式的组合核函数。使用多尺度数学形态学方法从主成分变换后的图像上提取空间特征，通过组合核函数融合图像的光谱特征与空间特征，采用组合核函数相关向量机进行分类。使用AVIRIS高光谱图像对算法进行了验证。实验结果表明，与传统的基于光谱特征的相关向量机分类器相比，组合核函数相关向量机方法的总体精度和Kappa系数均有明显提升。同时，组合核函数相关向量机能够用较少的训练样本获得较高的分类精度，在高光谱图像分类中具有实用价值。
Abstract: This paper presents a composite kernel Relevance Vector Machine (RVM) algorithm for hyperspec-tral image classification. This paper constructs three forms of composite kernels based on the properties of kernels. The spatial feature is extracted using multi-scale morphological method from the image after principal components transform. The final classification is achieved by our composite kernel RVM algorithm. The proposed approach is tested in experiments on AVIRIS data. Com-pared with spectral kernel RVM, the overall accuracy and Kappa coefficient of the composite kernel RVM algorithm increased obviously. However, the training time dose not increase. Meanwhile, composite kernel RVM has ability to get high accuracy with relative small raining set. The proposed method has practical use in hyperspectral imagery classification.
文章引用：孙琤, 刘东好. 基于组合核函数相关向量机的高光谱图像分类[J]. 计算机科学与应用, 2023, 13(7): 1399-1408. https://doi.org/10.12677/CSA.2023.137138

参考文献

[1] 杜培军, 夏俊士, 薛朝辉, 等. 高光谱遥感影像分类研究进展[J]. 北京遥感学报, 2016, 20(2): 236-256.
[2] 张锦水, 何春阳, 潘耀忠, 等. 基于SVM的多源信息复合的高空间分辨率遥感数据分类研究[J]. 遥感学报, 2006, 10(1): 49-57.
[3] Melgani, F. and Bruzzone, L. (2004) Classification of Hyperspectral Remote Sensing Images with Support Vector Machines. IEEE Transactions on Geoscience and Remote Sensing, 42, 1778-1790. [Google Scholar] [CrossRef] 
[4] 胡潭高, 潘耀忠, 张锦水, 李苓苓, 等. 基于线性光谱模型和支撑向量机的软硬分类方法[J]. 光谱学与光谱分析, 2011, 31(2): 508-511.
[5] Tarabalka, Y., Fauvel, M., Chanussot, J. and Benediktsson, J.A. (2010) SVM-and MRF-Based Method for Accurate Classification of Hyperspec-tral Images. IEEE Geoscience and Remote Sensing Letters, 7, 736-740. [Google Scholar] [CrossRef] 
[6] Ren, Y.M., Zhang, Y.N., Meng, Q.J. and Zhang, L. (2012) Hyperspectral Image Classification Based on Multiple Improved Particle Swarm Cooperative Optimization and SVM. Proceedings of the 21st International Conference on Pattern Recognition (ICPR2012), Tsukuba, 11-15 November 2012, 2274-2277.
[7] 钟燕飞, 良培, 李平湘. 基于多值免疫网络的多光谱遥感影像分类[J]. 计算机学报, 2007, 30(12): 2181-2188.
[8] 孔毅, 纪定哲, 程玉虎, 等. 基于光谱注意力图卷积网络的高光谱图像分类[J]. 电子与信息学报, 2023, 45(4): 1426-1434.
[9] 宋廷强, 宗达, 刘童心, 等. 联合分离卷积与密集连接轻量级神经网络的高光谱图像分类[J]. 遥感学报, 2022, 26(11): 2317-2328.
[10] Jimenez, L.O., Rivera-Medina, J.L., Rodríguez-Díaz, E., Arzuaga-Cruz, E. and Ramirez-Velez, M. (2005) Integration of Spatial and Spectral Information by Means of Unsu-pervised Extraction and Classification for Homogenous Objects Applied to Multispectral and Hyperspectraldata. IEEE Transactions on Geoscience and Remote Sensing, 43, 844-851. [Google Scholar] [CrossRef] 
[11] Benediktsson, J.A., Palmason, J.A. and Sveinsson, J.R. (2005) Classification of Hyperspectral Data from Urban Areas Based on Extended Morphological Profiles. IEEE Transactions on Geoscience and Remote Sensing, 43, 480-491. [Google Scholar] [CrossRef] 
[12] Benediktsson, J.A., Pesaresi, M. and Amason, K. (2003) Classi-fication and Feature Extraction for Remote Sensing Images from Urban Areas Based on Morphological Transformations. IEEE Transactions on Geoscience and Remote Sensing, 41, 1940-1949. [Google Scholar] [CrossRef] 
[13] Tuia, D., Ratle, F., Pozdnoukhov, A. and Camps-Valls, G. (2010) Multisource Composite Kernels for Urban-Image Classification. IEEE Geoscience and Remote Sensing Letters, 7, 88-92. [Google Scholar] [CrossRef] 
[14] Tipping, M.E. (2001) Sparse Bayesian Learning and the Rele-vance Vector Machine. The Journal of Machine Learning Research, 1, 211-244.
[15] 吴慧兰, 刘国栋, 浦昭邦. 基于相关向量机的惯性约束聚变实验靶识别技术研究[J]. 光学学报, 2009, 29(7): 1888-1891.
[16] Foody, G.M. (2008) RVM-Based Multi-Class Classification of Remotely Sensed Data. International Journal of Remote Sensing, 29, 1817-1823. [Google Scholar] [CrossRef] 
[17] Demir, B. and Erturk, S. (2007) Hyperspectral Im-age Classification Using Relevance Vector Machines. IEEE Geoscience and Remote Sensing Letters, 4, 586-590. [Google Scholar] [CrossRef] 
[18] Camps-Valls, G., Rodrigo-Gonzalez, A., Munoz-Mari, J., Gomez-Chova, L. and Calpe-Maravilla, J. (2007) Hyperspectral Image Classification with Mahalanobis Relevance Vector Machines. 2007 IEEE International Geoscience and Remote Sensing Symposium, Barcelona, 23-28 July 2007, 3802-3805. [Google Scholar] [CrossRef] 
[19] 赵春晖, 齐滨, 张燚. 基于改进型相关向量机的高光谱图像分类[J]. 光学学报, 2012, 32(8): 256-261.
[20] Tipping, M.E. and Faul, A. (2003) Fast Marginal Like-lihood Maximisation for Sparse Bayesian Models. Proceedings of the Ninth International Workshop on Artificial Intelli-gence and Statistics, Key West, 3-6 January 2003, 276-283.
[21] Vapnik, V.N. (1998) Statistical Learning Theory. Wiley, New York, 423.
[22] Camps-Valls, G., Gomez-Chova, L., Munoz-Mari, J., Vila-Frances, J. and Cal-pe-Maravilla, J. (2006) Composite Kernels for Hyperspectral Image Classification. IEEE Geoscience and Remote Sensing Letters, 3, 93-97. [Google Scholar] [CrossRef] 
[23] MultiSpec (2022) A Freeware Multispectral Image Data Analysis System. http://ftp.ecn.purdue.edu/biehl/MultiSpec/92AV3C/