A band selection method based on improved subspace partition
-
摘要: 高光谱图像具有光谱分辨率高、波段连续、数据量大、图谱合一等特点。然而较高的光谱分辨率会造成波段间相关性强,信息冗余多。所以如何从数百个高光谱波段中选出有利于识别或分类的波段组合成为了高光谱应用需要解决的问题。文章针对相邻波段间相关性较大的特点,提出一种改进的对波段相关矩阵进行全局搜索的子空间划分的波段选择方法。该方法克服了传统只利用相关向量对波段进行划分的缺陷,利用整个相关矩阵进行全局搜索划分,再在划分后的子空间内进行波段选择,从而降低了波段之间的相关性。文章最后使用上述方法对AVIRIS数据进行波段选择,并通过SVM方法对其进行地物分类,结果表明该方法较不进行子空间划分的波段选择方法有较高的分类精度。Abstract: Hyperspectral image has hundreds of successively narrow bands, which brings serious problems such as large correlation and redundant information. The selection of the optimal bands, which are suited for classification or recognition, has become a difficult work that needs to be overcome. In order to solve the problem of the large correlation among bands, a band selection method based on improved subspace partition through global search on correlation matrix was proposed. Through a global search, the band correlation matrix was divided into a series of subspace, from which the optimal bands were finally selected. The proposed method provides a band selection which has small correlation between each other. The result of an experiment which used Support Vector Machine(SVM) on an AVIRIS image shows that the proposed method is valid.
-
Key words:
- band selection /
- hyperspectral image /
- subspace partition
-
[1] Wang Jingli. Maximum band screening and its application to hyperspectral target detection[J]. Infrared and Laser Engineering, 2012, 41(6):1514-1519.(in Chinese) 王静荔.波段最大筛选法及其在高光谱目标探测中的应用[J]. 红外与激光工程, 2012, 41(6):1514-1519. [2] [3] Wang Yiting, Huang Shiqi, Liu Daizhi, et al. Novel band selection method based on target detection[J]. Infrared and Laser Engineering, 2013, 42(8):2294-2298.(in Chinese) 王艺婷,黄世奇,刘代志,等.一种新的基于目标检测的波段选择方法[J]. 红外与激光工程, 2013, 42(8):2294-2298. [4] [5] [6] Bajcsy P, Groves P. Methodology for hyper spectral band selection photogram metric engineering and remote sensing[J]. Photogrammetric Engineering and Remote Sensing Journal, 2004, 70(7):793-802. [7] Liu Jianping, Zhao Yingshi, Sun Shuling. Experimental studies about methods on optimal bands selection of hyperspectral remote sensing datasets[J]. Remote Sensing Technology and Application, 2001, 16(1):7-13.(in Chinese) 刘建平,赵英时,孙淑玲.高光谱遥感数据最佳波段选择方法试验研究[J]. 遥感技术与应用, 2001, 16(1):7-13. [8] [9] Yang Jinhong. Optimal band selection methods of hypersepctral remote sensing data[D]. Nanjing:Nanjing University of Information ScienceTechnology, 2005.(in Chinese) 杨金红.高光谱遥感数据最佳波段选择方法研究[D]. 南京:南京信息工程大学, 2005. [10] [11] [12] Liu Chunhong, Zhao Chunhui, Zhang Lingyan. A new method of hyperspectral remote sensing image dimensional reduction[J]. Journal of Image and Graphics, 2005, 10(2) :218-222.(in Chinese) 刘春红,赵春晖,张凌雁.一种新的高光谱遥感图像降维方法[J]. 中国图象图形学报, 2005, 10(2):218-222. [13] Gu Yanfeng, Zhang Ye. Feature extraction based on automatic subspace partition for hypersepctral images[J]. Remote Sensing Technology and Application, 2003, 18(6):32-35.(in Chinese) 谷延锋,张晔.基于自动子空间划分的高光谱数据特征提取[J]. 遥感技术与应用, 2003, 18(6):32-35. [14] [15] Dong Yanhua. Research on key technology of hyperspectral remote sensing image processing[D]. Harbin:Harbin University of Science and Technology, 2006.(in Chinese) 董延华.超光谱遥感图像处理关键技术研究[D]. 哈尔滨:哈尔滨理工大学, 2006. [16] [17] Han Ruimei, Yang Minhua. Study on an improved method of band selection of hypersepctral remote sensing data[J]. Geomatics Spatial Information Technology, 2010, 33(3):137-139.(in Chinese) 韩瑞梅,杨敏华.一种改进的高光谱遥感数据波段选择方法的研究[J]. 测绘与空间地理信息, 2010, 33(3):137-139. [18] [19] Guo Baofeng, Gunn Steve R, Damper R I. Band selection for hyperspectral image classification using mutual information[J]. IEEE Geoscience and Remote Sensing Letters, 2006, 3(4):522-526. [20] [21] Wang Liguo, Wei Fangjie. Band selection for hypersepctral imagery based on combination of genetic algorithm and ant colony algorithm[J]. Journal of Image and Graphics, 2013, 18(2):235-242.(in Chinese) 王立国,魏芳洁.结合遗传算法和蚁群算法的高光谱图像波段选择[J]. 中国图象图形学报, 2013, 18(2):235-242. -

计量
- 文章访问数: 301
- HTML全文浏览量: 55
- PDF下载量: 209
- 被引次数: 0