基于分类器集成的高光谱遥感图像分类方法

高光谱遥感图像为地物的精确分类带来了机遇，但也面临着一些挑战，高光谱遥感图像分类中所面临的一个挑战是如何处理高的光谱维数和小的样本数目之间的矛盾，目前几乎全部采用降维方法来缓解这一矛盾。集成学习的出现和选择性集成概念的提出为解决这一问题提供了新的研究思路，基于这一思想提出了基于波段分组和分类器集成的方法。在高光谱遥感图像的原始光谱空间根据波段之间的相似性信息对光谱波段进行分类，从每类中随机抽取一个波段形成新的光谱组，并依靠限制不同光谱组中相同波段的数目增加不同光谱组之间的差异程度，将新的光谱组作为训练分类器的特征子集，在特征子集训练最大似然分类器，使用简单的多数投票法合成得到最终的集成分类器。实验结果表明，使用基于波段分组和分类器集成的方法可以得到更高的分类精度。

Abstract

The conflict of high dimensionality and the very limited number of available training samples is one of the problems in the classification of hyperspectral images. At the same time, the redundance between different bands brings trouble to the classification. The ensemble learning provides a new way for solving the problem mentioned above. Based on the correlation between different bands, a band grouping is carried out. By selecting different bands from different groups new subsets of spectral bands is formed. The redundance reduces bands in the new spectral band subsets are independent and used to train the maximum likelihood (ML) classifiers which can be used later for ensembling. The combining of classifiers is done by the simple majority voting and the ensemble classifier is formed. Experimental results of the hyperspectral remotely sensing image demonstrate that the method presented here has an excellent classification result and outpeforms many other methods.

参考文献

[1] 宋琳, 程咏梅, 赵永强. 基于稀疏表示模型和自回归模型的高光谱分类[J]. 光学学报, 2012, 32(3): 0330003.

Song Lin, Cheng Yongmei, Zhao Yongqiang. Hyper-spectrum classification based on sparse representation model and auto-regressive model [J]. Acta Optica Sinica, 2012, 32(3): 0330003.

[2] 吴超, 吴一全. 基于混沌粒子群优化投影寻踪的高光谱图像目标检测[J]. 光学学报, 2011, 31(12): 1211003.

Wu Chao, Wu Yiquan. Target detection in hyperspectral image using projection pursuit based on chaotic particle swarm optimization [J]. Acta Optica Sinica, 2011, 31(12): 1211003.

[3] Paul M Mather, Magaly Koch. Computer Processing of Remotely-Sensed Images: An Introduction [M]. NewYork: John Wiley & Sons, 2011. 229-285.

[4] B M Shahshahani, D A Landgrebe. The effect of unlabeled samples in reducing the small sample size problem and mitigating the hughes phenomenon [J]. IEEE Transactions on Geoscience and Remote Sensing, 1994, 32(5): 1087-1095.

[5] 刘小刚, 赵慧洁, 李娜. 基于多重分形谱的高光谱数据特征提取[J]. 光学学报, 2009, 29(3): 844-847.

Liu Xiaogang, Zhao Huijie, Li Na. Feature extraction based on multifraetal spectrum for hyperspectral data [J]. Acta Optica Sinica, 2009, 29(3): 844-847.

[6] Q Jackson, D A Landgrebe. An adaptive classifier design for high-dimensional data analysis with a limited training dataset [J]. IEEE Transactions on Geoscience and Remote Sensing, 2001, 39(12): 2664-2679.

[7] L Zhang, B Du. Recent advances in hyperspectral image processing [J]. Geo-spatial Information Science, 2012, 15(3): 143-156.

[8] T G Dietterich. Machine learning research: four current directions [J]. Ai Magazine, 1997, 18(4): 97-136.

[9] 周杨, 厉小润, 赵辽英. 改进的高光谱图像线性预测波段选择算法[J]. 光学学报, 2013, 33(8): 0828002.

Zhou Yang, Li Xiaorun, Zhao Liaoying. Modified linear-prediction based band selection for hyperspectral image [J]. Acta Optica Sinica, 2013, 33(8): 0828002.

[10] 葛亮, 王斌, 张立明. 基于波段聚类的高光谱图像波段选择[J]. 计算机辅助设计与图形学学报, 2012, 24(11): 1447-1454.

Ge Liang, Wang Bin, Zhang Liming. Band selection based on band clustering for hyperspectral imagery [J]. Journal of Computer-Aided Design & Computer Graphics, 2012, 24(11): 1447-1454.

[11] B Guo, S R Gunn, R I Damper, et al.. Band selection for hyperspectral image classification using mutual information [J]. Geoscience and Remote Sensing Letters, 2006, 3(4): 522-526.

[12] H Z M Shafri, A Suhaili, S Mansor. The performance of maximum likelihood, spectral angle mapper, neural network and decision tree classifiers in hyperspectral image analysis [J]. Journal of Computer Science, 2007, 3(6): 419-423.

[13] A Tsymbal, S Puuronena, D W Pattersonb. Ensemble feature selection with the simple bayesian classification [J]. Information Fusion, 2003, 4(2): 87-100.

[14] 谢元澄. 分类器集成研究[D]. 南京: 南京理工大学, 2009.

Xie Yuancheng. Research on Classifier ensemble [D]. Nanjing: Nanjing University of Science and Technology, 2009.

[15] Z H Zhou, J X Wu, W Tang. Ensembling neural networks: many could be better than all [J]. Artificial Intelligence, 2002, 137(1-2): 239-263.

[16] G Brown, L I Kuncheva. “Good” and “Bad” Diversity in Majority Vote Ensembles[M]. Multiple Classifier Systems Berlin Heidelberg: Springer, 2010. 124-133.

[17] L I Kuncheva, C J Whitaker, C A Shipp, et al.. Is independence good for combining classifiers [C]. IEEE 2000. Proceedings. 15th International Conference on Pattern recognition, 2000, 2: 168-171.

[18] L I Kuncheva, C J Whitaker, C A Shipp, et al.. Limits on the majority vote accuracy in classifier fusion [J]. Pattern Analysis and Applications, 2003, 6(1): 22-31.

[19] Y Liu, X Yao. Ensemble learning via negative correlation [J]. Neural Networks, 1999,12(10): 1399-1404.

[20] 唐耀华, 高静怀, 包乾宗. 一种新的选择性支持向量机集成学习算法[J]. 西安交通大学学报, 2008, 42(10): 1221-1225.

Tang Yaohua, Gao Jinghuai, Bao Qianzong. Novel selective support vector machine ensemble learning algorithm [J]. Journel of Xi′an Jiaotong University, 2008, 42(10): 1221-1225.

[21] B Wu, C Chen, T M Kechadi, et al.. A comparative evaluation of filter-based feature selection methods for hyper-spectral band selection [J]. International Journal of Remote Sensing, 2013, 34(22): 7974-7990.

[22] 刘颖. 基于半监督集成支持向量机的土地覆盖遥感分类方法研究[D]. 合肥: 中国科技大学, 2013. 44-45.

Liu Ying. The Study of Semisupervised Ensembled Support Vector Machines for Land Cover Classification [D]. Hefei: University of Chinese Academy of Sciences, 2013. 44-45.

[23] R A Montserud, R Leamans. Comparing global vegetation maps with the kappa statistic [J]. Ecological Modeling,1992, 62(4): 275-293.

[24] A Galal, H Hasan, I F Iman. Learnable hyperspectral measures [J]. Egyptian Informatics Journal, 2012, 13(2): 85-94.

[25] S V Forero, J Angulo. Classification of hyperspectral images by tensor modeling and additive morphological decomposition [J]. Pattern Recognition, 2013, 46(2): 566-577.

[26] N Alajian, Y Bazi, F Melgani, et al.. Fusion of supervised and unsupervised learning for improved classification of hyperspectral images [J]. Information Sciences, 2012, 217(6): 39-55.

[27] L David. Hyperspectral image data analysisi [J]. Signal Processing Magazine, 2002, 19(1): 17-28.

[28] G Valls, D Tuia, L Bruzzone, et al.. Advances in hyperspectral image classification: earth monitoring with statistical learning methods [J]. Signal Processing Magazine, 2014, 31(1): 45-54.

樊利恒, 吕俊伟, 邓江生. 基于分类器集成的高光谱遥感图像分类方法[J]. 光学学报, 2014, 34(9): 0910002. Fan Liheng, Lü Junwei, Deng Jiangsheng. Classification of Hyperspectral Remote Sensing Images Based on Bands Grouping and Classification Ensembles[J]. Acta Optica Sinica, 2014, 34(9): 0910002.

基于分类器集成的高光谱遥感图像分类方法

关于本站 Cookie 的使用提示

全站搜索

基于分类器集成的高光谱遥感图像分类方法

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索