DOI: 
10.22389/0016-7126-2021-970-4-54-64
1 Yamashkin S.A.
2 Yamashkin A.A.
3 Zanozin V.V.
4 Barmin A.N.
Year: 
№: 
970
Pages: 
54-64

Ogarev Mordovia State University

1, 
2, 

Astrakhan State University

3, 
4, 
Abstract:
The authors propose their solving the task of improving the accuracy of remote sensing data classification under conditions of labeled data scarcity through using a geosystem approach that involves analyzing the genetic uniformity of various-scale territorially adjacent formations and hierarchical levels. The advantage of the proposed GeoSystemNet model is a great number of freedom degrees, which enables flexible configuration of the model based on the task being solved. Testing the GeoSystemNet model for classifying the EuroSAT set, algorithmically expanded from the perspective of the geosystem approach, showed the possibility of increasing the classification accuracy under the conditions of training data scarcity within 9 %, as well as approaching the accuracy of the deep ResNet50 and GoogleNet models. The authors note that the use of the geosystem approach according to the methodology proposed in the article for solving the above-mentioned problem requires an individual project approach to the formation of the data for analysis.
References: 
1.   Beruchashvili N.L. Chetyre izmereniya landshafta. Moskva: Mysl, 1986, 182 p.
2.   Kikin P. M., Kolesnikov A. A., Portnov A. M., Grishchenko D. V. Analiz i prognozirovanie prostranstvenno-vremennykh ekologicheskikh pokazatelei s ispol'zovaniem metodov mashinnogo obucheniya. InterKarto. InterGIS, 2020, Vol. 26, no. 3, pp. 53–61. DOI: 10.35595/2414-9179-2020-3-26-53-61.
3.   Nikolaev V. A. Landshaftovedenie. Moskva: Izd-vo MGU, 2000, 2000, 94 p.
4.   Sochava V.B. Vvedenie v uchenie o geosistemah. Novosibirsk: Nauka, 1978, 320 p.
5.   Bengio Y., LeCun Y. (2007) Scaling learning algorithms towards AI. Large-scale kernel machines, no. 34 (5), pp. 1-41.
6.   Boureau Y. L., Ponce J., LeCun Y. (2010) A theoretical analysis of feature pooling in visual recognition. Proceedings of the 27th international conference on machine learning (ICML-10). Haifa, Israel. pp. 111-118.
7.   Chollet F. (2017) Xception: Deep learning with depthwise separable convolutions. Proceedings of the IEEE conference on computer vision and pattern recognition. Honolulu, USA. pp. 1251-1258. DOI: 10.1109/CVPR.2017.195.
8.   Helber P., Bischke B., Dengel A., Borth D. (2019) Eurosat: A novel dataset and deep learning benchmark for land use and land cover classification. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, no. 12 (7), pp. 2217-2226. DOI: 10.1109/IGARSS.2018.8519248.
9.   Hu X., Weng Q. (2009) Estimating impervious surfaces from medium spatial resolution imagery using the self-organizing map and multi-layer perceptron neural networks. Remote Sensing of Environment, no. 113 (10), pp. 2089-2102. DOI: 10.1016/j.rse.2009.05.014.
10.   Ioffe S., Szegedy C. (2015) Batch normalization: Accelerating deep network training by reducing internal covariate shift. arXiv: 1502.03167 [cs. LG].
11.   Kim D. E., Gourbesville P., Liong S. Y. (2019) Overcoming data scarcity in flood hazard assessment using remote sensing and artificial neural network. Smart Water, no. 4 (1), 2 p. DOI: 10.1186/s40713-018-0014-5.
12.   Kussul N., Lavreniuk M., Skakun S., Shelestov A. (2017) Deep learning classification of land cover and crop types using remote sensing data. IEEE Geoscience and Remote Sensing Letters, no. 14 (5), pp. 778-782. DOI: 10.1109/LGRS.2017.2681128.
13.   LeCun Y., Bengio Y., Hinton G. (2015) Deep learning. Nature, no. 521 (7553), pp. 436. DOI: 10.1038/nature14539.
14.   Schowengerdt R. A. (2006) Remote sensing: models and methods for image processing, 3 ed. Orlando: Academic Press. pp. 387-456.
15.   Yamashkin S., Radovanovic M., Yamashkin A., Vukovic D. (2018) Using ensemble systems to study natural processes. Journal of Hydroinformatics, no. 20 (4), pp. 753-765. DOI: 10.2166/hydro.2018.076.
Citation:
Yamashkin S.A., 
Yamashkin A.A., 
Zanozin V.V., 
Barmin A.N., 
(2021) Development of an algorithm for the Earth remote sensing data classification using deep machine learning methods for analyzing the geosystem model of the territory. Geodesy and cartography = Geodezia i Kartografia, 82(4), pp. 54-64. (In Russian). DOI: 10.22389/0016-7126-2021-970-4-54-64
Publication History
Received: 06.07.2020
Accepted: 16.02.2021
Published: 20.05.2021

Content

2021 April DOI:
10.22389/0016-7126-2021-970-4

QR-code page

QR-код страницы