(47-5) 14 * << * >> * Russian * English * Content * All Issues

Ensembles of spectral-spatial convolutional neural network models for classifying soil types in hyperspectral images
N.A. Firsov 1,2, V.V. Podlipnov 1,2, N.A. Ivliev 1,2, D.D. Ryskova 2, A.V. Pirogov 1,2, A.A. Muzyka 1,2, A.R. Makarov 1,2, V.E. Lobanov 2,3, V.I. Platonov 2, A.N. Babichev 2, V.A. Monastyrskiy 2, V.I. Olgarenko 2, D.P. Nikolaev 4, R.V. Skidanov 1,2, A.V. Nikonorov 1,2, N.L. Kazanskiy 1,2, V.A. Soyfer 1,2

1 IPSI RAS – Branch of the FSRC "Crystallography and Photonics" RAS,
443001, Samara, Russia, Molodogvardeyskaya 151;
2 Samara National Research University, 443086, Samara, Russia, Moskovskoye Shosse 34;
3 Adyghe State University,
385000, Maykop, Republic of Adygea, Russia, Pervomayskaya St. 208;
4 Institute for Information Transmission Problems of the Russian Academy of Sciences (Kharkevich Institute),
127051, Moscow, Russia, Bol'shoi Karetnyi per. 19

 PDF, 3792 kB

DOI: 10.18287/2412-6179-CO-1260

Pages: 795-805.

Full text of article: Russian language.

Abstract:
The paper presents a study of various approaches to the classification of soil covers based on neural network algorithms using hyperspectral remote and proximal sensing of the Earth. The spectral distributions were recorded in the laboratory using an Offner imaging scanning hyperspectrometer. Spectral-spatial characteristics of nine soil samples from various parts of a farming land in the Samara region were experimentally studied. Using a method of energy dispersion microanalysis, the correspondence between the hyperspectral data and the chemical composition of the samples taken was established. Based on the data obtained, a neural network-aided classification of soil samples was implemented depending on the content of constituent elements such as carbon and calcium. A normalized spectral-spatial convolutional neural network was used as a classifier. As a result of the work, an approach to the classification of high-resolution hyper-spectral images based on the refinement of a multiclass convolutional neural network using an ensemble of binary classifiers is proposed. It is shown that the classification of soil samples by carbon and calcium content is carried out with an accuracy of 0.96.

Keywords:
hyperspectral images, hyperspectral sensing, proximal sensing, convolutional neural networks, spectral-spatial classification, soil cartography.

Citation:
Firsov NA, Podlipnov VV, Ivliev NA, Ryskova DD, Pirogov AV, Muzyka AA, Makarov AR, Lobanov VE, Platonov VI, Babichev AN, Monastyrskiy VA, Olgarenko VI, Nikolaev DP, Skidanov RV, Nikonorov AV, Kazanskiy NL, Soifer VA. Ensembles of spectral-spatial convolutional neural network models for classifying soil types in hyperspectral images. Computer Optics 2023; 47(5): 795-805. DOI: 10.18287/2412-6179-CO-1260.

Acknowledgements:
This work was partly funded under the ggovernemtn project of the FSRC "Crystallography and Photonics" RAS (experimental part) and grant No.20-69-47110 from the Russian Science Foundation (theoretical part).

References:
  1. Xue J, Su B. Significant remote sensing vegetation indices: A review of developments and applications. J Sens 2017; 2017: 1353691. DOI: 10.1155/2017/1353691.
  2. Li H, Wang C, Zhong C, Su A, Xiong C, Wang J, Liu J. Mapping urban bare land automatically from Landsat imagery with a simple index. Remote Sens 2017; 9: 249. DOI: 10.3390/rs9030249.
  3. Rasul A, Balzter H, Faqe Ibrahim GR, Hameed HM, Wheeler J, Adamu B, Ibrahim S, Najmaddin PM Applying built-up and bare-soil indi-ces from Landsat 8 to cities in dry climates. Land 2018; 7(3): 81. DOI: 10.3390/land7030081.
  4. Deng C, Wu C. BCI: A biophysical composition index for remote sensing of urban environments. Remote Sens Environ 2012; 127: 247-259. DOI: 10.1016/j.rse.2012.09.009.
  5. Zhao H, Chen X. Use of normalized difference bareness index in quickly mapping bare areas from TM/ETM+. Proceedings. 2005 IEEE Int Geoscience and Remote Sensing Symposium (IGARSS '05) 2005; 3: 1666-1668. DOI: 10.1109/IGARSS.2005.1526319.
  6. Schneider A, Friedl MA, Potere D. Mapping global urban areas using MODIS 500-m data: New methods and datasets based on ‘urban ecoregions’. Remote Sens Environ 2010; 114(8): 1733-1746. DOI: 10.1016/j.rse.2010.03.003.
  7. Hestir EL, Greenberg JA, Ustin SL. Classification trees for aquatic vegetation community prediction from imaging spectroscopy. IEEE J Sel Top Appl Earth Obs Remote Sens 2012; 5(5): 1572-1584. DOI: 10.1109/JSTARS.2012.2200878.
  8. Hu X, Weng Q. Estimating impervious surfaces from medium spatial resolution imagery using the self-organizing map and multi-layer perceptron neural networks. Remote Sens Environ 2009; 113(10): 2089-2102. DOI: 10.1016/j.rse.2009.05.014.
  9. Zhang F, Yang X. Improving land cover classification in an urbanized coastal area by random forests: The role of variable selection. Remote Sens Environ 2020; 251: 112105. DOI: 10.1016/j.rse.2020.112105.
  10. Li W. Mapping urban impervious surfaces by using spectral mixture analysis and spectral indices. Remote Sens 2019; 12(1): 94. DOI: 10.3390/rs12010094.
  11. Somers B, Asner G, Tits L, Coppin P. Endmember variability in Spectral Mixture Analysis: A review. Remote Sens Environ 2011; 115(7): 1603-1616. DOI: 10.1016/j.rse.2011.03.003.
  12. Estoque R, Murayama Y. Classification and change detection of built-up lands from Landsat-7 ETM+ and Landsat-8 OLI/TIRS imageries: A comparative assessment of various spectral indices. Ecol Indic 2015; 56: 205-217. DOI: 10.1016/j.ecolind.2015.03.037.
  13. Becker F, Choudhury BJ. Relative sensitivity of normalized difference vegetation index (NDVI) and microwave polarization difference index (MPDI) for vegetation and desertification monitoring. Remote Sens Environ 1988; 24(2): 297-311. DOI: 10.1016/0034-4257(88)90031-4.
  14. McFeeters SK. The use of the normalized difference water index (NDWI) in the delineation of open water features. Int J Remote Sens 1996; 17: 1425-1432. DOI: 10.1080/01431169608948714.
  15. Xu HQ. Analysis of impervious surface and its impact on urban heat environment using the normalized difference impervious surface index (NDISI). Photogramm Eng Remote Sensing 2010; 76(5): 557-565. DOI: 10.14358/PERS.76.5.557.
  16. DengYB, Wu CS, Li M, Chen RR. RNDSI: A ratio normalized difference soil index for remote sensing of urban/suburban environments. Int J Appl Earth Obs Geoinf 2015; 39: 40-48.
  17. Palacios-Orueta A, Ustin SL. Remote sensing of soil properties in the Santa Monica Mountains I. Spectral analysis. Remote Sens Environ 1998; 65(2): 170-183.
  18. Rikimaru A, Roy PS, Miyatake S. Tropical forest cover density mapping. Trop Ecol 2002; 43: 39-47.
  19. Liu Y, Meng Q, Zhang L, Wu C. NDBSI: A normalized difference bare soil index for remote sensing to improve bare soil mapping accuracy in urban and rural areas. Catena 2022; 214: 106265.
  20. Zhu Y, Wang D, Zhang H, Shi P. Soil organic carbon content retrieved by UAV-borne high resolution spectrometer. Transactions of the Chinese Society of Agricultural Engineering 2021; 37(6): 66-72.
  21. Zhou Q, Ding J, Tang M, Yang B. Inversion of soil organic matter content in oasis typical of arid area and its influencing factors. Acta Pedologica Sinica 2018; 55(2): 313-324.
  22. Hamza MM, Blank VA, Podlipnov VV, Doskolovich LL, Skidanov RV, Fan B. Spectral lenses to highlight blood vessels in the skin. Computer Optics 2022; 46(6): 899-904. DOI: 10.18287/2412-6179-CO-1155.
  23. Wei H, Yangyu H, Li W, Fan Zh, Hengchao L. Deep convolutional neural networks for hyperspectral image classification. J Sens 2015; 2015: 258619. DOI: 10.1155/2015/258619.
  24. He M, Li B, Chen H. Multi-scale 3D deep convolutional neural network for hyperspectral image classification. IEEE Int Conf on Image Processing (ICIP) 2017: 3904-3908. DOI: 10.1109/ICIP.2017.8297014.
  25. Firsov N, Podlipnov V, Ivliev N, Nikolaev P, Mashkov S, Ishkin P, Skidanov R, Nikonorov A. Neural network-aided classification of hyperspectral vegetation images with a training sample generated using an adaptive vegetation index. Computer Optics 2021; 45(6): 887-896. DOI: 10.18287/2412-6179-CO-1038.
  26. Makarov AR, Podlipnov VV, Ivliev NA, Nikonorov AV, Ulyanov DI, Firsov NA. Neural network classification of coffee varieties on hyperspectral images. VIII Int Conf on Information Technology and Nanotechnology (ITNT) 2022: 1-3. DOI: 10.1109/ITNT55410.2022.9848735.

© 2009, IPSI RAS
151, Molodogvardeiskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru ; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846) 332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20