Comparison analysis of spatial and spectral feature in vegetation classification based on AVIRIS hyperspectral image

doi:10.12133/j.smartag.2020.2.1.201911-SA005

Abstract

Abstract:

With the development of hyperspectral sensor technology and remote sensing data acquisition platform, the application of hyperspectral data is becoming more and more popular in precision agriculture. Spectral features and spatial features are two main kinds of features used in hyperspectral image classification. The comparison of spectral features and spatial features in vegetation classification of hyperspectral image is a special application in hyperspectral image classification. Therefore, this study compared the performance of several typical spectral features and spatial features in vegetation classification of hyperspectral image. The considered spatial features include grey level co-occurrence matrix (GLCM) based features, Gabor features and morphological features. The considered spectral feature selection or extraction methods include minimal-redundancy-maximal-relevance (mRMR), joint mutual information (JMI), conditional mutual information maximization (CMIM), double input symmetrical relevance (DISR), Jeffreys-Matusita (JM), principal component analysis (PCA), independent component analysis (ICA) and linear discriminant analysis (LDA). PCA, an effective subspace feature extraction method, is widely used in the feature extraction of hyperspectral image. The first several principal components (PCs) are usually selected as spectral features in hyperspectral image classification. However, the first several PCs have no guarantee to achieve good class separability and classification accuracy. Considering that, a hybrid feature extraction approach named as PCA_ScatterMatrix was proposed which combined PCA and an improved scatter-matrix-based feature selection method, aiming to select PCs with high class separability and get high overall classification accuracy. The experiments and comparative analyses were conducted with a widely used hyperspectral image, which was collected over the agricultural area in northwestern Indiana, USA (United States of America) by the AVIRIS (Airborne Visible / Infrared Imaging Spectrometer). The experimental results indicated that: (1) The proposed hybrid feature extraction method PCA_ScatterMatrix got the highest overall classification accuracy on both data sets (82.7% and 86.5%) among three classic subspace feature extraction methods (PCA, ICA and LDA) and respectively improved overall classification accuracy by 1.5% and 2.5% on both data sets, comparing to original PCA; (2) Compared to spectral features, spatial feature extraction methods generally got higher overall classification accuracy, especially Gabor spatial features got the highest overall classification accuracy on both data sets (95.5% and 96.7%). The results suggest that the proposed method is effective in vegetation classification of hyperspectral image and the spatial features play a much more important role in vegetation classification of hyperspectral image, comparing with spectral features.

Key words: hyperspectral image, vegetation classification, spectral feature, spatial feature, hybrid feature extraction method, scatter-matrix, principal component analysis

CLC Number:

S127

Fu Yuanyuan, Yang Guijun, Duan Dandan, Zhang Yongtao, Gu Xiaohe, Yang Xiaodong, Xu Xingang, Li Zhenhai. Comparison analysis of spatial and spectral feature in vegetation classification based on AVIRIS hyperspectral image[J]. Smart Agriculture, 2020, 2(1): 68-76.

Figures/Tables 6

Fig. 1

Fig. 2

Table 1

Table 2

Table 3

Table 4

References 31

1	Yang G, Liu J, Zhao C, et al. Unmanned aerial vehicle remote sensing for field-based crop phenotyping: Current status and perspectives[J]. Front Plant Sci, 2017, 8: 1111.
2	陈仲新，郝鹏宇，刘佳，等. 农业遥感卫星发展现状及我国监测需求分析[J]. 智慧农业, 2019,1(1): 32-42.
	Chen Z, Hao P, Liu J, et al. Technical demands for agricultural remote sensing satellites in China[J]. Smart Agriculture, 2019, 1(1): 32-42.
3	Li Z, Li Z, Fairbairn D, et al. Multi-LUTs method for canopy nitrogen density estimation in winter wheat by field and UAV hyperspectral[J]. Computers and Electronics in Agriculture, 2019, 162: 174-182.
4	Xu Y, Zhang L， Du B, et al. Spectral-spatial unified networks for hyperspectral image classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, (99): 1-17.
5	Sun W， Du Q. Hyperspectral band selection: A review[J]. IEEE Geoscience and Remote Sensing Magazine, 2019, 7(2): 118-139.
6	Fu Y, hao C, Wang J, et al. An improved combination of spectral and spatial features for vegetation classification in hyperspectral images[J]. Remote Sensing, 2017, 9(3): no. 261.
7	付元元. 基于遥感数据的作物长势参数反演及作物管理分区研究[D]. 杭州：浙江大学， 2015.
	Fu Y. Remote sensing data based crop growth parameters retrieval and crop management zone delineation research [D]. Hangzhou: Zhejiang University, 2015.
8	夏道平，付元元，王纪华，等. 分散矩阵特征选择方法改进及在高光谱影像植被分类中的应用[J]. 农业工程学报， 2016， 32(21): 196-201.
	Xia D, Fu Y, Wang J, et al. An improved scatter-matrix-based feature selection method for vegetation classification of hyperspectral image[J]. Transactions of the CSAE, 2016, 32(21): 196-201.
9	Cgerutadat A, Bruce L M. Why hyperspectral component analysis is not an appropriate feature extraction method for hyperspectral data[J]. IEEE International Geoscience and Remote Sensing Symposium, 2003, 6: 3420-3422.
10	Uddin M P, Mamun M A, Hossain M A. Effective feature extraction through segmentation-based folded-PCA for hyperspectral image classification[J]. International Journal of Remote Sensing, 2019， 40: 1-31.
11	Haralick R M, Shanmugam K, Dinstein I. Textural features for image classification[J]. Studies in Media and Communication, 1973, SMC-3(6): 610-621.
12	Jia S, Deng X, Zhu J, et al. Collaborative representation-based multiscale superpixel fusion for hyperspectral image classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(10)： 7770-7784.
13	Benediktsson J A, Palmason J A, Seveinsson J R. Classification of hyperspectral data from urban areas based on extended morphological profiles[J]. IEEE Transactions on Geoscience and Remote Sensing. 2005, 43(3)： 480-491.
14	王增茂，杜博，张良培，等. 基于纹理特征和形态特征融合的高光谱影像分类法[J]. 光子学报， 2014, 43(8):122-129.
	Wang Z, Du B, Zhang L， et al. Based on texture feature and extend morphological profile fusion for hyperspectral image classification[J]. Acta Photonica Sinca, 2014, 43(8): 122-129.
15	Jia S, Wu K, Zhu J, et al. Spectral-spatial Gabor surface feature fusion approach for hyperspectral imagery classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 57(2): 1142-1154.
16	Lv F, Han M. Hyperspectral image classification based on multiple reduced kernel extreme learning machine[J]. International Journal of Machine Learning and Cybernetics, 2019, 10: 3397-3405.
17	Xue Z. A general generative adversarial capsule network for hyperspectral image spectral-spatial classification[J]. Remote Sensing Letters, 2020, 11(1): 19-28.
18	Huang H, Duan Y, He H, et al. Spatial-spectral local discriminant projection for dimensionality reduction of hyperspectral image[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2019, 156： 77-93.
19	Li J, Huang X, Gamba P, et al. Multiple feature learning for hyperspectral image classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(3): 1592-606.
20	Chang C, Lin C. LIBSVM: A library for support vector machines[J]. ACM Transactions on Intelligent Systems and Technology (TIST), 2011, 2(3): 27.
21	Peng H, Long F, Ding C. Feature selection based on mutual information criteria of max-dependency, max-relevance, and min-redundancy[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(8)： 1226-1238.
22	Brown G, Pocock A, Zhao M J, et al. Conditional likelihood maximisation: A unifying framework for information theoretic feature selection[J]. Journal of Machine Learning Research, 2012, 13: 27-66.
23	Meyer P E, Schretter C, Bontempi G. Information-theoretic feature selection in microarray data using variable complementarity[J]. IEEE Journal of Selected Topics in Signal Processing, 2008, 2(3): 261-274.
24	Sun K, Geng X, Ji L. A new sparsity-based band selection method for target detection of hyperspectral image[J]. IEEE Geoscience and Remote Sensing Letters, 2015, 12(2): 329-333.
25	Wang J, Chang C. Independent component analysis-based dimensionality reduction with applications in hyperspectral image analysis[J]. IEEE Transactions on Geoscience and Remote Sensing, 2006, 44(6): 1586-1600.
26	Du Q. Unsupervised real-time constrained linear discriminant analysis to hyperspectral image classification[J]. Pattern Recognition, 2007, 40: 1510-1519.
27	Zhou L, Wang L, Shen C. Feature selection with redundancy-constrained class separability[J]. IEEE Transactions on Neural Networks. 2010, 21(5): 853-858.
28	Huang X, Liu X, Zhang L. A multichannel gray level co-occurrence matrix for multi/hyperspectral image texture representation[J]. Remote Sensing， 2014, 6(9): 8424-8445.
29	Jiang L, Narayanan R M. Integrated spectral and spatial information mining in remote sensing imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2004, 42(3): 673-685.
30	Manjunath B S, Ma W Y. Texture features for browsing and retrieval of image data[J]. IEEE Transaction on Pattern Analysis and Machine Intelligence, 1996, 18(8): 837-842.
31	Zhang L, Zhang L, Tao D, et al. A multifeature tensor for remote-sensing target recognition[J]. IEEE Geoscience and Remote Sensing Letters, 2011, 8(2): 374-378.

分类方法	特征数目	总体分类正确率	Kappa系数	生产者精度
分类方法	特征数目	总体分类正确率	Kappa系数	玉米	大豆	小麦	森林	草地
PCA	6	0.840	0.764	0.807	0.824	0.982	0.932	0.712
ICA	15	0.827	0.746	0.822	0.822	1.000	0.843	0.792
PCA_ScatterMatrix	13	0.865	0.799	0.820	0.878	0.982	0.909	0.746
LDA	4	0.824	0.743	0.815	0.784	0.982	0.905	0.856

方法	波段数目	总体正确率	Kappa系数	生产者精度
方法	波段数目	总体正确率	Kappa系数	玉米	大豆	玉米	森林	草地
JMI	33	0.778	0.847	0.846	0.830	0.990	0.904	0.703
mRMR	40	0.725	0.603	0.632	0.742	0.905	0.813	0.591
CMIM	32	0.837	0.764	0.791	0.836	0.990	0.900	0.752
DISR	40	0.858	0.795	0.860	0.849	0.981	0.882	0.780
JM	19	0.803	0.718	0.808	0.758	0.971	0.882	0.773
GLCM	8	0.857	0.798	0.864	0.803	1.000	0.915	0.990
Gabor	80	0.955	0.935	0.980	0.918	1.000	0.986	1.000
Morph	20	0.904	0.860	0.870	0.884	0.990	0.967	0.941

方法	波段数目	总体正确率	Kappa系数	生产者精度
方法	波段数目	总体正确率	Kappa系数	玉米	大豆	小麦	森林	草地
JMI	40	0.863	0.798	0.823	0.861	0.982	0.913	0.839
mRMR	40	0.757	0.644	0.708	0.732	0.945	0.883	0.623
CMIM	32	0.846	0.773	0.784	0.844	0.982	0.917	0.839
DISR	23	0.844	0.768	0.760	0.852	0.982	0.938	0.750
JM	24	0.856	0.788	0.843	0.843	1.000	0.899	0.818
GLCM	8	0.874	0.818	0.894	0.825	1.000	0.921	0.987
Gabor	80	0.967	0.951	0.984	0.944	1.000	0.987	1.000
Morph	20	0.926	0.890	0.871	0.928	1.000	0.970	0.987

[1]	ZHU Chao, WU Fan, LIU Changbin, ZHAO Jianxiang, LIN Lili, TIAN Xueying, MIAO Teng. Tassel Segmentation of Maize Point Cloud Based on Super Voxels Clustering and Local Features [J]. Smart Agriculture, 2021, 3(1): 75-85.
[2]	FENG Qingchun, CHEN Jian, CHENG Wei, WANG Xiu. Multi-Band Image Fusion Method for Visually Identifying Tomato Plant’s Organs With Similar Color [J]. Smart Agriculture, 2020, 2(2): 126-134.
[3]	Yang Chenghai. Airborne remote sensing systems for precision agriculture applications [J]. Smart Agriculture, 2020, 2(1): 1-22.