Please wait a minute...

Frontiers of Earth Science

Front. Earth Sci.    2015, Vol. 9 Issue (2) : 225-236     DOI: 10.1007/s11707-014-0473-4
RESEARCH ARTICLE |
Hyperspectral image classification based on volumetric texture and dimensionality reduction
Hongjun SU1,2,Yehua SHENG3,Peijun DU2,*(),Chen CHEN4,Kui LIU4
1. School of Earth Sciences and Engineering, Hohai University, Nanjing 210098, China
2. Key Laboratory of Satellite Mapping Technology and Application, National Administration of Surveying, Mapping and Geoinformation, Nanjing University, Nanjing 210046, China
3. Key Laboratory of Virtual Geographic Environment (Ministry of Education), Nanjing Normal University, Nanjing 210023, China
4. Department of Electrical Engineering, the University of Texas at Dallas, Richardson, TX 75080-3021, USA
Download: PDF(7184 KB)   HTML
Export: BibTeX | EndNote | Reference Manager | ProCite | RefWorks
Abstract

A novel approach using volumetric texture and reduced-spectral features is presented for hyperspectral image classification. Using this approach, the volumetric textural features were extracted by volumetric gray-level co-occurrence matrices (VGLCM). The spectral features were extracted by minimum estimated abundance covariance (MEAC) and linear prediction (LP)-based band selection, and a semi-supervised k-means (SKM) clustering method with deleting the worst cluster (SKMd) band-clustering algorithms. Moreover, four feature combination schemes were designed for hyperspectral image classification by using spectral and textural features. It has been proven that the proposed method using VGLCM outperforms the gray-level co-occurrence matrices (GLCM) method, and the experimental results indicate that the combination of spectral information with volumetric textural features leads to an improved classification performance in hyperspectral imagery.

Keywords fusion      hyperspectral imagery      image classification      volumetric textural feature      spectral feature     
Corresponding Authors: Peijun DU   
Online First Date: 14 November 2014    Issue Date: 30 April 2015
 Cite this article:   
Hongjun SU,Yehua SHENG,Peijun DU, et al. Hyperspectral image classification based on volumetric texture and dimensionality reduction[J]. Front. Earth Sci., 2015, 9(2): 225-236.
 URL:  
http://journal.hep.com.cn/fesci/EN/10.1007/s11707-014-0473-4
http://journal.hep.com.cn/fesci/EN/Y2015/V9/I2/225
Service
E-mail this article
E-mail Alert
RSS
Articles by authors
Hongjun SU
Yehua SHENG
Peijun DU
Chen CHEN
Kui LIU
Fig.1  Hyperspectral image cube for Washington DC Mall data
Fig.2  Texture computation of GLCM and VGLCM
Directions(θ, ψ) Vector(dx, dy, dz) Symmetrical directions(θ, ψ) Vector(dx, dy, dz)
1 (0°, 0°) (0, 0, 1) (180°, 0°) (0, 0, -1)
2 (0°, 45°) (1, 0, 1) (180°, 135°) (-1, 0, -1)
3 (0°, 90°) (1, 0, 0) (180°, 90°) (-1, 0, 0)
4 (0°, 135°) (1, 0, -1) (180°, 45°) (-1, 0, 1)
5 (45°, 45°) (1, 1, 1) (225°, 135°) (-1, -1, -1)
6 (45°, 90°) (1, 1, 0) (225°, 90°) (-1, -1, 0)
7 (45°, 135°) (1, 1, -1) (225°, 45°) (-1, -1, 1)
8 (90°, 45°) (0, 1, 1) (270°, 135°) (0, -1, -1)
9 (90°, 90°) (0, 1, 0) (270°, 90°) (0, -1, 0)
10 (90°, 135°) (0, 1, -1) (270°, 45°) (0, -1, 1)
11 (135°, 45°) (-1, 1, 1) (315°, 135°) (1, -1, -1)
12 (135°, 90°) (-1, 1, 0) (315°, 90°) (1, -1, 0)
13 (135°, 135°) (-1, 1, -1) (315°, 45°) (1, -1, 1)
Tab.1  Directions of pairwise pixels for VGLCM
Fig.3  Two-dimensional (2D; left) and three-dimensional (3D; right) semi-variogram analysis for texture extraction
Fig.4  Classification framework for hyperspectral image analysis
Classification with spectral data Classification with spectral and texture
Scheme I All bands All bands+texture features
Scheme II PCA PCA+texture features
Scheme III Selected bands Selected bands+texture features
Scheme IV Band clusters Band clusters+texture features
Tab.2  Feature fusion schemes for hyperspectral image classification
Class name Training No. Test No.
Road 55 892
Grass 57 910
Shadow 50 567
Trail 46 623
Tree 49 656
Roof 52 1123
Total 309 4771
Tab.3  Training and test samples for DC Mall data
No. Class names All ground truth in our data
1 Alfalfa 54
2 Corn-notill 1434
3 Corn-min 834
4 Corn 234
5 Grass/pasture 497
6 Grass/trees 747
7 Grass/pasture-mowed 26
8 Hay-windrowed 489
9 Oats 20
10 Soybeans-notill 968
11 Soybeans-min 2468
12 Soybeans-clean 614
13 Wheat 212
14 Woods 1294
15 Bldg-grass-trees 380
16 Stone-steel-towers 95
Total samples 21025
Tab.4  Training and test samples for Indian Pines
Fig.5  Texture samples for different objects
Fig.6  Various box sizes for VGLCM of D.C. Mall dataset. (a) Road I; (b) Road II; (c) Trees I; (d) Trees II; (e) Grass I; (f) Grass II; (g) Building I; (h) Building II
Fig.7  Various window sizes for GLCM of D.C. Mall dataset
Fig.8  Textures extracted using GLCM (left) and VGLCM (right). (a) Variance; (b) Contrast; (c) Dissimilarity; (d) Energy; (e) Entropy; (f) Homogeneity
Fig.9  Classification maps using GLCM and VGLCM textures for D.C. Mall data. (a) 5 selected bands+6 texture features (GLCM); (b) 5 selected bands+6 texture features (VGLCM); (c) 5 clusters+6 texture features (GLCM); (d) 5 clusters+6 texture features (VGLCM)
OA All bands+ PCA 5 PCs+ 5 Selected bands+ 5 Clusters+
GLCM VGLCM GLCM VGLCM GLCM VGLCM GLCM VGLCM
Variance 0.9187 0.9126 0.9365 0.9304 0.9329 0.9449 0.9319 0.9541
Contrast 0.9342 0.9333 0.9357 0.9354 0.9476 0.9466 0.9568 0.9558
Dissimilarity 0.9233 0.8979 0.9287 0.9401 0.9430 0.9466 0.9587 0.9610
Energy 0.9191 0.9113 0.9310 0.9346 0.9317 0.9468 0.9407 0.9595
Entropy 0.9166 0.9122 0.9329 0.9300 0.9231 0.9449 0.9310 0.9554
Homogeneity 0.9088 0.9057 0.9346 0.9466 0.9141 0.9338 0.9235 0.9426
6 texture features 0.9338 0.9396 0.8965 0.8991 0.8900 0.9185 0.8791 0.9103
Tab.5  
OA GLCM VGLCM
+ All bands+ +PCA 15 PCs +15 Selected bands +15 Clusters + All bands+ +PCA 15 +15 Selected bands +15 Clusters
Spectral data 0.8462 0.8289 0.8466 0.8873 0.8462 0.8289 0.8466 0.8873
Variance 0.8526 0.8324 0.8575 0.8934 0.8546 0.8384 0.8573 0.8880
Contrast 0.8535 0.8330 0.8508 0.8904 0.8532 0.8317 0.8539 0.8863
Dissimilarity 0.8535 0.8340 0.8532 0.8944 0.8555 0.8347 0.8555 0.8854
Energy 0.8518 0.8324 0.8569 0.8941 0.8522 0.8376 0.8588 0.8871
Entropy 0.8532 0.8321 0.8579 0.8922 0.8545 0.8362 0.8567 0.8879
Homogeneity 0.8536 0.8311 0.8582 0.8939 0.8580 0.8370 0.8540 0.8897
6 texture features 0.8644 0.8475 0.8646 0.8942 0.8732 0.8677 0.8836 0.8962
Tab.6  Classification results for spectral and texture feature fusion
Fig.10  Classification comparison using spectral and texture data
1 Angelo N P, Haertel V (2003). On the application of Gabor filtering in supervised image classification. Int J Remote Sens, 24(10): 2167–2189
doi: 10.1080/01431160210163146
2 Benediktsson J A, Palmason J A, Sveinsson J R (2005). Classification of hyperspectral data from urban areas based on extended morphological profiles. IEEE Trans Geosci Rem Sens, 43(3): 480–491
doi: 10.1109/TGRS.2004.842478
3 Bernabe S, Marpu P R, Plaza A, Mura M D, Benediktsson J A (2014). Spectral-spatial classification of multispectral images using kernel feature space representation. IEEE Geosci Remote Sens Lett, 11(1): 288–292
doi: 10.1109/LGRS.2013.2256336
4 Chang C I (2003). Hyperspectral Imaging: Techniques for Spectral Detection and Classification. New York: Kluwer Academic/Plenum Publishers, 13–15
5 Chang C I (2013). Hyperspectral Data Processing: Algorithm Design and Analysis. New Jersey: Wiley-Interscience, 1–5
6 Chen C, Li W, Tramel E W, Cui M, Prasad S, Fowler J E (2014a). Spectral–spatial preprocessing using multihypothesis prediction for noise-robust hyperspectral image classification. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 7(4): 1047–1059
doi: 10.1109/JSTARS.2013.2295610
7 Chen C, Li W, Tramel E W, Fowler J E (2014b). Reconstruction of hyperspectral imagery from random projections using multihypothesis prediction. IEEE Trans Geosci Rem Sens, 52(1): 365–374
doi: 10.1109/TGRS.2013.2240307
8 Gamba P, Dell’ Acqua F, Lisini G, Trianni G (2007). Improved VHR urban area mapping exploiting object boundaries. IEEE Trans Geosci Rem Sens, 45(8): 2676–2682
doi: 10.1109/TGRS.2007.899811
9 Haralick R M, Shanmugam K, Dinstein I H (1973). Texture features for image classification. IEEE Trans Syst Man Cybern, 3(6): 610–621
doi: 10.1109/TSMC.1973.4309314
10 Huang X, Zhang L, Gong W (2011). Information fusion of aerial images and LIDAR data in urban areas: vector stacking, re-classification, and post-processing approaches. Int J Remote Sens, 32(1): 69–84
doi: 10.1080/01431160903439882
11 Jackson Q, Landgrebe D (2002). Adaptive bayesian contextual classification based on markov random fields. IEEE Trans Geosci Rem Sens, 40(11): 2454–2463
doi: 10.1109/TGRS.2002.805087
12 Li J, Bioucas-Dias J, Plaza A (2012). Spectral-spatial hyperspectral image segmentation using subspace multinomial logistic regression and markov random fields. IEEE Trans Geosci Rem Sens, 50(3): 809–823
doi: 10.1109/TGRS.2011.2162649
13 Liu K, Du Q, Yang H, Ma B (2010). Optical flow and principle component analysis-based motion detection in outdoor videos. EURASIP J Adv Signal Process, 2010: 680623
doi: 10.1155/2010/680623
14 Liu K, Ma B, Du Q, Chen G (2012). Fast motion detection from airborne videos using graphics processing units. J Appl Remote Sens, 6(1): 061505
15 Marceau D J, Howarth P J, Dubois J M, Gratton D J (1990). Evaluation of the grey-level co-occurrence matrix method, for land-cover classification using SPOT imagery. IEEE Trans Geosci Rem Sens, 28(4): 513–519
doi: 10.1109/TGRS.1990.572937
16 Mokji M M, Bakar S A R A (2007). Adaptive thresholding based on co-occurrence matrix edge information. Journal of Computers, 2(8): 44–52
doi: 10.4304/jcp.2.8.44-52
17 Mura M D, Villa A, Benediktsson J A, Chanussot J, Bruzzone L (2011). Classification of hyperspectral images by using extended morphological attribute profiles and independent component analysis. IEEE Geosci Remote Sens Lett, 8(3): 541–545
18 Nyoungui A N, Tonye E, Akono A (2002). Evaluation of speckle filtering and texture analysis methods for land cover classification from SAR images. Int J Remote Sens, 23(9): 1895–1925
doi: 10.1080/01431160110036157
19 Plaza J, Plaza A, Barra C (2009). Multi-channel morphological profiles for classification of hyperspectral image data using support vector machines. Sensors (Basel Switzerland), 9(1): 196–218
doi: 10.3390/s90100196
20 Rahman A F, Gamon J A, Sims D A, Schmidts M (2003). Optimum pixel size for hyperspectral studies of ecosystem function in southern California chaparral and grassland. Remote Sens Environ, 84(2): 192–207
doi: 10.1016/S0034-4257(02)00107-4
21 Rajadell O, Garcia-Sevilla P, Pla F (2013). Spectral–spatial pixel characterization using gabor filters for hyperspectral image classification. IEEE Geosci Remote Sens Lett, 10(4): 860–864
doi: 10.1109/LGRS.2012.2226426
22 Su H, Du Q (2012). Hyperspectral band clustering and band selection for urban land cover classification. Geocarto Int, 27(5): 395–411
doi: 10.1080/10106049.2011.643322
23 Su H, Sheng Y, Du P, Liu K (2012). Adaptive affinity propagation with spectral angle mapper for semi-supervised hyperspectral band selection. Appl Opt, 51(14): 2656–2663
doi: 10.1364/AO.51.002656
24 Su H, Yang H, Du Q, Sheng Y (2011). Semi-supervised band clustering for dimensionality reduction of hyperspectral imagery. IEEE Geosci Remote Sens Lett, 8(6): 1135–1139
doi: 10.1109/LGRS.2011.2158185
25 Tsai F, Chang C K, Rau J Y, Lin T H, Liu G R (2007). 3D computation of gray level co-occurrence in hyperspectral image cubes. Lect Notes Comput Sci, 4679: 429–440
doi: 10.1007/978-3-540-74198-5_33
26 Yang H, Du Q, Su H, Sheng Y (2011). An efficient method for supervised hyperspectral band selection. IEEE Geosci Remote Sens Lett, 8(1): 138–142
doi: 10.1109/LGRS.2010.2053516
27 Yang H, Ma B, Du Q, Yang C (2010). Improving urban land use and land cover classification from high-spatial-resolution hyperspectral imagery using contextual information. J Appl Remote Sens, 4(1): 041890
doi: 10.1117/1.3491192
Related articles from Frontiers Journals
[1] Sizhang TANG,Chaomin SHEN,Guixu ZHANG. Adaptive regularized scheme for remote sensing image fusion[J]. Front. Earth Sci., 2016, 10(2): 236-244.
[2] Shaolei TANG,Xiaofeng YANG,Di DONG,Ziwei LI. Merging daily sea surface temperature data from multiple satellites using a Bayesian maximum entropy method[J]. Front. Earth Sci., 2015, 9(4): 722-731.
[3] Wei WANG,Xinfeng YAO,Minhe JI,Jiao ZHANG. Spectral data treatments for impervious endmember derivation and fraction mapping from Landsat ETM+ imagery: a comparative analysis[J]. Front. Earth Sci., 2015, 9(2): 179-191.
[4] Zhan TIAN, Honglin ZHONG, Runhe SHI, Laixiang SUN, Günther FISCHER, Zhuoran LIANG. Estimating potential yield of wheat production in China based on cross-scale data-model fusion[J]. Front Earth Sci, 2012, 6(4): 364-372.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed