To tackle the low classification accuracy problem of hyperspectral image classification caused by similar samples under the condition of small-sample-size, this paper proposes a hyperspectral image classification algorithm based on Fisher criterion and TrAdboost (H_TrAdaboost). Firstly, an auxiliary sample is determined with the spectral angle mapping (SAM) method and spectral information divergence (SID) so as to improve the total number of training samples. Secondly, the samples are studied separability based on the improved Fisher criterion to obtain relatively strong samples set. Finally, the weight of positive and negative sample distribution is adjusted dynamically by using TrAdaboost algorithm so as to achieve hyperspectral similarity class classification in the small sample size problem. In the comparative experiments with other compared algorithms, the highest classification accuracy is achieved, which fully shows that H_TrAdaboost algorithm can well solve the similar hyperspectral image classifications.
Du P J, Xia J S, Xue Z H , et al. Review of hyperspectral remote sensing image classification[J]. Journal of Remote Sensing, 2016,20(2):236-256.
[2]
Camps-Valls G, Tuia D, Bruzzone L , et al. Advances in hyperspectral image classification:Earth monitoring with statistical learning methods[J]. IEEE Signal Processing Magazine, 2014,31(1):45-54.
doi: 10.1109/MSP.2013.2279179
[3]
高恒振 . 高光谱遥感图像分类技术研究[D]. 长沙:国防科学技术大学, 2011.
Gao H Z . Research on Classification Techinique for Hyperspectral Remote Sensing Imagery[D]. Changsha:National University of Defence Technology, 2011.
[4]
谭琨 . 基于支持向量机的高光谱遥感影像分类研究[D]. 徐州:中国矿业大学, 2010.
Tan K . Research on Hyperspectral Romote Sensing Image Classification Based on Support Vector Machine[D]. Xuzhou:China University of Mining Technology, 2010.
Yang K G, Feng X Z, Xiao P F , et al. Optimal subspace ensemble with SVM for hyperspectral image classification[J]. Journal of Remote Sensing, 2016,20(3):409-419.
[6]
Persello C, Bruzzone L . Active and semisupervised learning for the classification of remote sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2014,52(11):6937-6956.
doi: 10.1109/TGRS.2014.2305805
[7]
Pan S J, Yang Q . A survey on transfer learning[J]. IEEE Transactions on Knowledge and Data Engineering, 2010,22(10):1345-1359.
doi: 10.1109/TKDE.2009.191
Wu T J, Luo J C, Xia L G , et al. An automatic sample collection method for object-oriented classification of remotely sensed imageries based on transfer learning[J]. Acta Geodaetica et Cartographica Sinica, 2014,43(9):908-916.
[9]
Zhou Y, Lian J, Han M . Remote sensing image transfer classification based on weighted extreme learning machine[J]. IEEE Geoscience and Remote Sensing Letters, 2016,13(10):1405-1409.
doi: 10.1109/LGRS.2016.2568263
[10]
Dai W, Yang Q, Xue G R, et al. Boosting for transfer learning [C]// Proceedings of the 24th International Conference on Machine Learning.Corvallis:ACM, 2007: 193-200.
Xia B, Zhao X J . An unsupervised band selection algorithm based on spectral information divergence[J]. Computer Applications and Software. 2015,32(7):212-215,251.
Liu W J, Yang X H, Qu H C , et al. Hyperspectral unmixing algorithm based on spectral information divergence and spectral angle mapping[J]. Journal of Computer Application, 2015,35(3):844-848.
Wang H R, Xia K W, Ren M M , et al. Adaptive denoising method of hyperspectral remote sensing based on PCA and dictionary learning[J]. Journal of Computer Application, 2016,36(12):3411-3417.
Xu M M, Zhang L P, Du B , et al. Supervised band selection based on class separability for hyperspectral image[J]. Computer Science, 2015,42(4):274-275,296.
Wu D Y, Ma L . Multi-manifold LE algorithm for dimension reduction and classification of multitemporal hyperspectral image[J]. Remote Sensing for Land and Resources, 2018,30(2):80-86.doi: 10.6046/gtzyyg.2018.02.11.
Zhao B, Su H J, Cai Y . A hyperspectral image classification method based on collaborative representation in tangent space[J]. Geomatics and Information Science of Wuhan University, 2018,43(4):555-562.
2018, Vol. 30 
), 曲海成
