Please wait a minute...
国土资源遥感  2019, Vol. 31 Issue (2): 17-23    DOI: 10.6046/gtzyyg.2019.02.03
  技术方法 本期目录 | 过刊浏览 | 高级检索 |
加权空-谱主成分分析的高光谱图像分类
阿茹罕1,何芳2,王标标3
1.西安培华学院会计与金融学院,西安 710065
2.火箭军工程大学核工程学院,西安 710025
3.96862部队,洛阳 471003
Hyperspectral images classification via weighted spatial-spectral dimensionality reduction principle component analysis
Ruhan A1,Fang HE2,Biaobiao WANG3
1.Xi’an Peihua University, School of Accounting and Finance, Xi’an 710065, China;
2.Rocket Force Engineering University, School of Nuclear Engineering, Xi’an 710025, China;
3.Troops No.96862, Luoyang 471003, China
全文: PDF(9503 KB)   HTML  
输出: BibTeX | EndNote (RIS)      
摘要 

为了提高高光谱图像的分类精度,有效利用高光谱图像的空间信息和光谱信息对高光谱图像进行预处理,提出了一种新的空-谱联合降维方法——加权空-谱主成分分析(weighted spatial spectral principle component analysis,WSSPCA)算法。该算法结合高光谱图像的物理特性对高光谱图像进行重构,平滑了高光谱图像中存在的奇异点干扰; 然后,采用主成分分析(principle component analysis,PCA)方法对重构后的图像进行降维,降低了波段间的冗余性,有利于后续分类。在2组常用的高光谱数据集PaviaU和Indian Pines上进行实验结果表明,训练样本随机选取每一类地物的5%和10%的情况下,由WSSPCA算法得到的Kappa系数最大值分别达到了0.955 9和0.896 1,较基准线分别提高了0.193 8和0.205 0,分类结果明显优于其他算法。

服务
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章
阿茹罕
何芳
王标标
关键词 高光谱图像分类加权空-谱主成分分析降维    
Abstract

In order to improve the hyperspectral images (HSI) classification accuracy and preprocess HSI by fully using the spatial and spectral information, this paper proposes a new spatial-spectral dimensionality reduction method, i.e., weighted spatial and spectral principle component analysis (WSSPCA). This algorithm reconstructs the HSI by using the physical characteristics of HSI, which can lower the influence of singular point in HSI. Principle component analysis (PCA) is utilized to reduce dimensionality of HIS, and it reduces the redundancy between bands and improves the HSI classification accuracy efficiently. The benchmark tests on PaviaU and Indian Pines demonstrate that the performance of WSSPCA is better than that of PCA and LPP when 5% and 10% samples in each class (10 samples are chosen when the total samples in every class is less than 100) are chosen randomly as train samples. The best values of Kappa coefficient obtained by WSSPCA are 0.955 9 and 0.896 1 respectively on the HSI datasets, exceeding the baseline by 0.193 8 and 0.205 0.

Key wordshyperspectral images classification    weighted spatial and spectral principle component analysis    dimensionality reduction
收稿日期: 2018-01-16      出版日期: 2019-05-23
ZTFLH:  TP751  
基金资助:国家自然科学基金项目“快速稳定统一主次子空间跟踪算法研究”资助(61401471)
作者简介: 阿茹罕(1989-),女,博士研究生,讲师,主要从事机器学习方法研究及其在遥感图像处理中的应用。Email: aruhan890309@163.com。
引用本文:   
阿茹罕,何芳,王标标. 加权空-谱主成分分析的高光谱图像分类[J]. 国土资源遥感, 2019, 31(2): 17-23.
Ruhan A,Fang HE,Biaobiao WANG. Hyperspectral images classification via weighted spatial-spectral dimensionality reduction principle component analysis. Remote Sensing for Land & Resources, 2019, 31(2): 17-23.
链接本文:  
http://www.gtzyyg.com/CN/10.6046/gtzyyg.2019.02.03      或      http://www.gtzyyg.com/CN/Y2019/V31/I2/17
Fig.1  不同情况下像元点的近邻空间
Fig.2  PaviaU和Indian Pines图像及其地物类型
Fig.3  在PaviaU数据集上不同参数对分类精度的影响
Fig.4  在PaviaU数据集上的分类精度评价指标与维数的关系
算法 OA/%(维数) Kappa(维数)
基准线 82.33(1) 0.762 1(1)
PCA 82.35(20) 0.762 5(16)
LPP 82.42(5) 0.764 8(5)
WSSPCA 96.69(20) 0.955 9(21)
Tab.1  在PaviaU数据集上由各种算法得到的最高评价指标及其对应的维数
Fig.5  分类结果
Fig.6  在Indian Pines数据集上不同参数对分类精度的影响
Fig.7  在Indian Pines数据集上的分类精度评价指标与维数的关系
算法 OA/%(维数) Kappa(维数)
基准线 72.97(1) 0.691 1(1)
PCA 72.94(20) 0.690 8(20)
LPP 72.91(3) 0.682 9(3)
WSSPCA 90.90(21) 0.896 1(21)
Tab.2  在Indian Pines数据集上由各种算法得到的最高评价指标及其对应的维数
Fig.8  分类结果
[1] 黄鸿, 杨媚, 张满菊 . 基于稀疏鉴别嵌入的高光谱遥感影像分类[J]. 光学精密工程, 2013,21(11):2922-2930.
doi: 10.3788/OPE.20132111.2922
Huang H, Yang M, Zhang M J . Hyperspectral remote sensing image classification based on SDE[J]. Optics and Precision Engineering, 2013,21(11):2922-2930.
doi: 10.3788/OPE.20132111.2922
[2] 杨艺, 韩崇昭, 韩德强 . 利用特征子空间评价与多分类器融合的高光谱图像分类[J]. 西安交通大学学报, 2010,44(8):21-24.
Yang Y, Han C Z, Han D Q . Hyperspectral image classification based on feature subspace evaluation and multiple classification fusion[J]. Journal of Xi’an Jiaotong University, 2010,44(8):21-24.
[3] 粘永健, 苏令华, 孙蕾 , 等. 基于聚类的高光谱图像无损压缩[J]. 电子与信息学报, 2009,31(6):1271-1274.
doi: 10.3724/SP.J.1146.2008.00660
Nian Y J, Su L H, Sun L , et al. Lossless coding for hyperspectral image based on spectral cluster[J]. Journal of Electronics and Information Technology, 2009,31(6):1271-1274.
doi: 10.3724/SP.J.1146.2008.00660
[4] 常威威, 郭雷, 刘坤 , 等. 基于Contourlet变换和主成分分析的高光谱数据噪声消除方法[J]. 电子与信息学报, 2009,31(12):2892-2896.
doi: 10.3724/SP.J.1146.2008.01675
Chang W W, Guo L, Liu K , et al. Denoising of hyperspectral data based on Contourlet transform and principal component analysis[J]. Journal of Electronics and Information Technology, 2009,31(12):2892-2896.
doi: 10.3724/SP.J.1146.2008.01675
[5] 刘嘉敏, 罗甫林, 黄鸿 , 等. 应用相关近邻局部线性嵌入算法的高光谱遥感影像分类[J]. 光学精密工程, 2014,22(6):1668-1676.
Liu J M, Luo F L, Huang H , et al. Classification of hyperspectral remote sensing images using correlation neighbor LLE[J]. Optics and Precision Engineering, 2014,22(6):1668-1676.
[6] 何同弟 . 高光谱图像的分类技术研究[D]. 重庆:重庆大学, 2014.
He T D . The Classification Technology Research Based on Hyperspectral Image[D]. Chongqing:Chongqing University, 2014.
[7] 潘宗序, 禹晶, 肖创柏 , 等. 基于光谱相似性的高光谱图像超分辨率算法[J]. 自动化学报, 2014,40(12):2797-2807.
doi: 10.3724/SP.J.1004.2014.02797
Pan Z X, Yu J, Xiao C B , et al. Spectral similarity-based super resolution for hyperspectral images[J]. Acta Automatica Sinica, 2014,40(12):2797-2807.
doi: 10.3724/SP.J.1004.2014.02797
[8] Zhou Y C, Peng J T , Chen C L P .Dimension reduction using spatial and spectral regularized local discriminant embedding for hyperspectral image classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015,53(2):1082-1095.
doi: 10.1109/TGRS.2014.2333539
[9] 唐中奇, 付光远, 陈进 , 等. 基于多尺度分割的高光谱图像稀疏表示与分类[J]. 光学精密工程, 2015,23(9):2708-2714.
Tang Z Q, Fu G Y, Chen J , et al. Multiscale segmentation-based sparse coding for hyperspectral image classification[J]. Optics and Precision Engineering, 2015,23(9):2708-2714.
[10] He R, Hu B G, Zheng W S , et al. Robust principal component analysis based on maximum correntropy criterion[J]. IEEE Transactions on Image Processing, 2011,20(6):1485-1494.
doi: 10.1109/TIP.2010.2103949
[11] Zhou Y C, Peng J T , Chen C L P .Dimension reduction using spatial and spectral regularized local discriminant embedding for hyperspectral image classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015,53(2):1082-1095.
doi: 10.1109/TGRS.2014.2333539
[12] 金鹏磊 . 空-谱联合高光谱数据降维与分类方法研究[D]. 西安:西安电子科技大学, 2014.
Jing P L . Researches on Spatial-Spectral Based Dimensionality Reduction and Classification of Hypespectral Data[D]. Xi’an:Xidian University, 2014
[13] Bandos T V, Bruzzone L , Camps-Valls . Classification of hyperspectral image with regularized linear discriminant analysis[J]. IEEE Transactions on Geoscience and Remote Sensing, 2009,47(3):862-873.
doi: 10.1109/TGRS.2008.2005729
[14] Wang R, Nie F P, Yang X J , et al. Robust 2DPCA with non-greedy -norm maximization for image analysis[J]. IEEE Transactions on Cybernetics, 2015,45(5):1108-1112.
[15] Lu X, Li X . Multiresolution imaging[J]. IEEE Transactions on Cybernetics, 2014,44(1):149-160.
doi: 10.1109/TCYB.2013.2286496
[16] He X F, Niyogi P. Locality preserving projections [C]//Advances in Neural Information Processing Systems(NIPS).Vancouver:MIT, 2004,16:153-160.
[17] Zhang Z Y, Zha H Y . Principal manifolds and nonlinear dimensionality reduction via local tangent space alignment[J]. Journal of Shanghai University, 2004,8(4):406-424
doi: 10.1007/s11741-004-0051-1
[18] 黄鸿, 郑新磊 . 加权空-谱与最近邻分类器相结合的高光谱图像分类[J]. 光学精密工程, 2016,24(4):873-881.
Huang H, Zheng X L . Hyperspectral image classification with combination of weighted spatial-spectral and KNN[J]. Optics and Precision Engineering, 2016,24(4):873-881.
[19] 何芳, 王榕, 于强 , 等. 加权空谱局部保持投影的高光谱图像特征提取[J]. 光学精密工程, 2017,25(1):263-273.
He F, Wang R, Yu Q , et al. Feature extraction of hyperspectral images of weighted spatial and spectral locality preserving projection (WSSLPP)[J]. Optics and Precision Engineering, 2017,25(1):263-273.
[20] Huang H, Luo F L, Liu J M , et al. Dimensionality reduction of hyperspectral images based on sparse discriminant manifold embedding[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2015,106:42-54.
doi: 10.1016/j.isprsjprs.2015.04.015
[1] 刘万军,李天慧,曲海成. 基于Fisher准则和TrAdaboost的高光谱相似样本分类算法[J]. 国土资源遥感, 2018, 30(4): 41-48.
[2] 孙小芳. 结合目标分割的高光谱城市地物分类[J]. 国土资源遥感, 2017, 29(3): 171-175.
[3] 陈洁, 杜磊, 李京, 韩亚超, 高子弘. 基于噪声白化的高光谱数据子空间维数算法[J]. 国土资源遥感, 2017, 29(2): 60-66.
[4] 樊雪, 刘清旺, 谭炳香. 基于机载PHI高光谱数据的森林优势树种分类研究[J]. 国土资源遥感, 2017, 29(2): 110-116.
[5] 鲁锦涛, 马丽. 基于流形对齐的高光谱遥感图像降维和分类算法[J]. 国土资源遥感, 2017, 29(1): 104-109.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
京ICP备05055290号-2
版权所有 © 2015 《国土资源遥感》编辑部
地址:北京学院路31号中国国土资源航空物探遥感中心 邮编:100083
电话:010-62060291/62060292 Email:gtzyyg@agrs.cn; gtzyyg@163.com
本系统由北京玛格泰克科技发展有限公司设计开发