Please wait a minute...
 
Remote Sensing for Land & Resources    2019, Vol. 31 Issue (3) : 59-64     DOI: 10.6046/gtzyyg.2019.03.08
|
Spatial and spectral feature hierarchical fusion for hyperspectral image feature extraction
Benzuo YAO1, Fang HE2
1. Anhui Police College, Hefei 230088, China
2. School of Nuclear Engineering,Rocket Force Engineering University, Xi’an 710025, China;
Download: PDF(3769 KB)   HTML
Export: BibTeX | EndNote | Reference Manager | ProCite | RefWorks    
Abstract  

In this paper, the multi-dimensional adaptive weighted filter (AWF) is used to filter the hyperspectral image with a certain dimension which are reduced by the feature extraction method based on spectral dimension. Then, the filter results obtained on all scales are hierarchical fusion into a new image, and the hierarchical fusion framework is designed. These treatments make the essential spatial and spectral features in hyperspectral images extracted effectively, so the classification accuracy is improved. The principal component analysis (PCA) algorithm is integrated into the framework, and a hierarchical fusion-principal component analysis (HF-PCA) algorithm is proposed. This method not only reduces the redundancy between bands, but also weakens the internal differences of the samples and improves the classification accuracy of hyperspectral images. Experimental results on the Indian Pines and Salinas databases demonstrate that the classification accuracy obtained by the HF-PCA algorithm is significantly higher than that of other algorithms, even when the number of training samples is small, and the maximum value of the overall classification accuracy is 86.73% and 95.01%, respectively. The classification accuracy of hyperspectral images is improved effectively.

Keywords spatial and spectral feature      hierarchical fusion      hierarchical fusion-principal component analysis      hyperspectral image classification     
:  TP751  
Issue Date: 30 August 2019
Service
E-mail this article
E-mail Alert
RSS
Articles by authors
Benzuo YAO
Fang HE
Cite this article:   
Benzuo YAO,Fang HE. Spatial and spectral feature hierarchical fusion for hyperspectral image feature extraction[J]. Remote Sensing for Land & Resources, 2019, 31(3): 59-64.
URL:  
https://www.gtzyyg.com/EN/10.6046/gtzyyg.2019.03.08     OR     https://www.gtzyyg.com/EN/Y2019/V31/I3/59
Fig.1  Adaptive weighted filters
Fig.2  Frame of HF-PCA
Fig.3  Indian Pines hyperspectral image
Fig.4  Salinas hyperspectral image
Fig.5  Classification results of different algorithms in Indian Pines dataset
地物类别 KNN PCA HF-PCA 地物类别 KNN PCA HF-PCA
Alfalfa 80.56 80.56 97.78 Oats 80.00 70.00 100
Corn-notill 46.57 45.84 84.32 Soybeans-notill 61.21 60.46 87.06
Corn-mintill 57.49 56.60 79.06 Soybeans-mintill 66.21 64.71 90.21
Corn 28.00 29.33 47.20 Soybeans-clean 44.40 43.16 70.05
Grass/pasture 78.87 78.21 87.93 Wheat 94.36 93.33 98.15
Grass/trees 94.37 94.52 97.63 Woods 88.02 88.69 95.97
Grass/pasture-mowed 100 100 97.78 Buildings-grass-tree-drives 41.14 39.78 62.94
Hay-windrowed 94.71 94.05 98.37 Stone-steel-towers 85.54 85.54 92.53
Tab.1  Classification accuracy of different types of features in Indian Pines dataset by different algorithms(%)
Fig.6  Classification results of different algorithms in Salinas dataset
地物类别 KNN PCA HF-PCA 地物类别 KNN PCA HF-PCA
Brocoli-green-weeds_1 98.54 98.54 99.75 Soil-vinyard-develop 97.49 97.39 99.56
Brocoli-green-weeds_2 98.54 98.43 99.70 Corn-senesced-green-weeds 83.94 83.20 92.49
Fallow 80.88 79.09 96.62 Lettuce-romained-4wk 88.93 88.93 89.76
Fallow-rough-plow 98.70 98.70 99.07 Lettuce-romained-5wk 100 100 99.97
Fallow-smooth 95.36 96.27 97.83 Lettuce-romained-6wk 97.57 96.91 99.05
Stubble 98.85 98.85 99.78 Lettuce-romained-7wk 87.54 87.44 96.83
Celery 98.93 98.93 99.60 Vinyard_untrained 59.65 59.32 88.25
Grapes-untrained 70.39 69.30 89.57 Vinyard_vertical-trellis 89.66 89.60 96.18
Tab.2  Classification accuracy of different types of features in Salinas dataset by different algorithms(%)
[1] 高晓惠 . 高光谱数据处理技术研究[D]. 西安:中国科学院大学, 2013.
[1] Gao X H . Research on Technologies of Hyperspectral Data Processing[D]. Xi’an:University of Chinese Academy of Science, 2013.
[2] Plaza A, Benediktsson J A, Boardman J W , et al. Recent advances in techniques for hyperspectral image processing[J]. Remote Sensing of Environment, 2009,113(s1):S110-S122.
[3] Bioucas-Dias J M, Plaza A, Camps-Valls G , et al. Hyperspectral remote sensing data analysis and future challenges[J]. IEEE Geoscience and Remote Sensing Magazine, 2013,1(2):6-36.
[4] 王晓飞, 侯传龙, 阎秋静 , 等. 基于相关向量机的高光谱图像噪声评估算法[J]. 红外与激光工程, 2014,43(12):4159-4163.
[4] Wang X F, Hou C L, Yan Q J , et al. Noise estimation algorithm based on relevance vector machine for hyperspectral imagery[J]. Infrared and Laser Engineering, 2014,43(12):4159-4163.
[5] 张成业, 秦其明, 陈理 , 等. 高光谱遥感岩矿识别的研究进展[J]. 光学精密工程, 2015,23(8):2407-2418.
[5] Zhang C Y, Qin Q M, Chen L , et al. Research and development of mineral identification utilizing hyperspectral remote sensing[J]. Optics and Precision Engineering, 2015,23(8):2407-2418.
[6] 邓承志, 张绍泉, 汪胜前 , 等. L1稀疏正则化的高光谱混合像元分解算法比较[J]. 红外与激光工程, 2015,44(3):1092-1097.
url: http://www.opticsjournal.net/Articles/Abstract?aid=OJ160126000272HdJgMj
[6] Deng C Z, Zhang S Q, Wang S Q , et al. Hyperspectral unmixing algorithm based on L1 regularization[J]. Infrared and Laser Engineering, 2015,44(3):1092-1097.
[7] 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.
[7] 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.
[8] 王琳, 李迅, 包云轩 , 等. 遥感技术在交通气象灾害监测中的应用进展[J]. 国土资源遥感, 2018,30(4):1-7.doi: 10.6046/gtzyyg.2018.04.01.
[8] Wang L, Li X, Bao Y X , et al. Research progress of remote sensing application meteorological disasters[J]. Remote Sensing for Land and Resources, 2018,30(4):1-7.doi: 10.6046/gtzyyg.2018.04.01.
[9] 叶珍 . 高光谱图像特征提取与分类算法研究[D]. 西安:西北工业大学, 2015.
[9] Ye Z . Research on Feature Extraction and Classification Algorithm for Hyperspectral Image[D]. Xi’an:Northwestern Polytechnical University, 2015.
[10] 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
[10] 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.
[11] 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.
[11] 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.
[12] Li H, Wei Y, Li L , et al. Hierarchical feature extraction with local neural response for image recognition[J]. IEEE Transactions on Systems Man and Cybernetics Part B:Cybernetics, 2012,43(2):412-424.
[13] Pu H Y, Chen Z H, Wang B . A novel spatial-spectral similarity measure for dimensionality reduction and classification of hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2014,52(11):7008-7022.
[14] 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.
[14] Jin P L . Researches on Spatial-Spectral Based Dimensionality Reduction and Classification of Hyperspectral Data[D]. Xi’an:Xidian University, 2014.
[1] HAN Yanling, CUI Pengxia, YANG Shuhu, LIU Yekun, WANG Jing, ZHANG Yun. Classification of hyperspectral image based on feature fusion of residual network[J]. Remote Sensing for Land & Resources, 2021, 33(2): 11-19.
[2] Wanjun LIU, Tianhui LI, Haicheng QU. Hyperspectral similar sample classification algorithm based on Fisher criterion and TrAdaboost[J]. Remote Sensing for Land & Resources, 2018, 30(4): 41-48.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
京ICP备05055290号-2
Copyright © 2017 Remote Sensing for Natural Resources
Support by Beijing Magtech