Synergistic use of PRISMA hyperspectral and Sentinel-1B SAR data for land cover classification
DOI:
https://doi.org/10.5564/mjgg.v62i46.4080Keywords:
Deep learning, Machine learning, Hyperspectral, SAR, Image classification stockAbstract
The aim of this study is to compare the performance of deep learning, machine learning, and advanced hyperspectral image classification methods for distinguishing land cover types in Ulaanbaatar city. The study area includes various land cover classes such as built-up areas, ger districts, forests, willows, grasslands, soil, and water, with significant statistical overlaps between the built-up areas and the ger districts. For data sources, we selected PRISMA (Hyperspectral Precursor of the Application Mission) and Sentinel-1B dual-polarization synthetic aperture radar (SAR) images. Three different band combinations were utilized to identify the mixed urban land cover classes in Mongolia's capital city. To differentiate the existing classes, we employed an artificial neural network (ANN), support vector machine (SVM), and spectral angle mapper (SAM), assessing their performance against one another. To evaluate the accuracy of the classification results, we applied the Kappa coefficient. For all three band combinations, the SVM method demonstrated superior performance, with Kappa coefficients ranging from 0.96 to 0.98. The ANN showed the second-highest performance, with Kappa coefficients ranging from 0.83 to 0.96. In contrast, the SAM yielded the lowest performance, with Kappa coefficients between 0.67 and 0.71. Our study observed that the performance of the selected classification techniques depended on the chosen parameters and the structure of the datasets. Overall, this study highlights that the combined use of hyperspectral and microwave datasets can enhance the classification of land cover types, with the SVM approach emerging as the most reliable method for producing an accurate land cover map.
Downloads
137
References
[1] D. Amarsaikhan, A. Enkhmanlai, V. Battsengel, and V. Batsaikhan, "Feature extraction and classification of hyperspectral images", CD-ROM Proceedings of the Asian Conference on Remote Sensing (RS), Pattaya, Thailand, 2012.
[2] A. Munkh-Erdene, D. Amarsaikhan, D. Enkhjargal, G. Odontuya, and E. Jargaldalai, “Feature Extraction Approach in Hyperspectral Data BT - Proceedings of the Environmental Science and Technology International Conference (ESTIC 2021),” Atlantis Press, 2021, pp. 102–108, Available: doi: 10.2991/aer.k.211029.019.
[3] G. Tejasree and L. Agilandeeswari, “An extensive review of hyperspectral image classification and prediction: techniques and challenges,” Multimed. Tools Appl., vol. 83, no. 34, pp. 80941–81038, 2024, Available: doi: 10.1007/s11042-024-18562-9.
[4] M. Ahmad, A. M. Khan, M. Mazzara, S. Distefano, M. Ali and M. S. Sarfraz, "A Fast and Compact 3-D CNN for Hyperspectral Image Classification," IEEE Geoscience and Remote Sensing Letters, vol. 19, pp. 1-5, 2022, Art no. 5502205, Available: doi: 10.1109/LGRS.2020.3043710.
[5] A. Patel, D. Vyas, N. Chaudhari, R. Patel, K. Patel, and D. Mehta, “Novel approach for the LULC change detection using GIS & Google Earth Engine through spatiotemporal analysis to evaluate the urbanization growth of Ahmedabad city,” Results Eng., vol. 21, p. 101788, Mar. 2024, Available: doi: 10.1016/j.rineng.2024.101788.
[6] D. Enkhjargal, D. Amarsaikhan, V. Battsengel, J. Sod-Erdene, and G. Tsogzol, "Applications of multitemporal optical images for forest resources study in Mongolia", Full paper published in CD-ROM Proceedings of the Asian Conference on RS, Nay Pyi Taw, Myanmar, 2014.
[7] J.Xie, J. Hua, S. Chen, P. Wu, P. Gao, D. Sun, Z. Lyu, S. Lyu, X. Xue, J. Lu, "HyperSFormer: A Transformer-Based End-to-End Hyperspectral Image Classification Method for Crop Classification", Remote Sens. 2023, 15, 3491, Available: doi:10.3390/rs15143491.
[8] M. F. Guerri, C. Distante, P. Spagnolo, F. Bougourzi, and A. Taleb-Ahmed, “Deep learning techniques for hyperspectral image analysis in agriculture: A review,” ISPRS Open J. Photogramm. Remote Sens., vol. 12, no. March, p. 100062, 2024, Available: doi: 10.1016/j.ophoto.2024.100062.
[9] Q. Shenming, L. Xiang, and G. Zhihua, “A new hyperspectral image classification method based on spatial-spectral features,” Sci. Rep., vol. 12, no. 1, pp. 1–16, 2022, Available: doi: 10.1038/s41598-022-05422-5.
[10] D. Amarsaikhan, A. Enkhmanlai, Ts. Bat-Erdene, E. Jargaldalai, and Ch. Bolorchuluun, "Feature extraction and classification of hyperspectral data of Mongolia using machine learning methods", Full paper published in eProceedings of the Asian Conference on RS, Ulaanbaatar, Mongolia, 2022.
[11] E. Amarsaikhan, D. Enkhjargal, E. Jargaldalai, and D. Amarsaikhan, “Comparison of machine learning and parametric methods for the discrimination of urban land cover types,” Geocarto Int., vol. 39, Jan. 2024, Available: doi: 10.1080/10106049.2024.2380372.
[12] PRISMA (Hyperspectral), [Online]. Available: www.eoportal.org/satellite-missions/ prisma-hyperspectral, 2023.
[13] A. Malekian and N. Chitsaz, "Chapter 4 - Concepts, procedures, and applications of artificial neural network models in streamflow forecasting", Editor(s): Priyanka Sharma, Deepesh Machiwal, Advances in Streamflow Forecasting, Elsevier, Pages 115-147, 2021, Available: doi:10.1016/B978-0-12-820673-7.00003-2.
[14] V. F. Rodriguez-Galiano and M. Chica-Rivas, “Evaluation of different machine learning methods for land cover mapping of a Mediterranean area using multi-seasonal Landsat images and Digital Terrain Models,” Int. J. Digit. Earth, vol. 7, no. 6, pp. 492–509, 2014, Available: doi: 10.1080/17538947.2012.748848.
[15] V. Sharma et al., “Insights into the recent advances of agro-industrial waste valorization for sustainable biogas production,” Elsevier, vol. Volume 390, p. 129829, 2023, Available: doi:10.1016/j.biortech.2023.129829.
[16] Y. Zhang, X. Sun, S. G. Bajwa, S. Sivarajan, J. Nowatzki, and M. Khan, “Chapter Nine - Plant Disease Monitoring With Vibrational Spectroscopy,” Elsevier, vol. 80, pp. 227–251, 2018, Available: doi: 10.1016/bs.coac.2018.03.006.
[17] S. Rashmi, S. Addamani, Venkat, and S. Ravikiran, “Spectal Angle Mapper Algorithm for remote Sensing Image Classification,” Int. J. Innov. Sci. Eng. Technol., vol. 1, no. 4, pp. 201–205, 2014.
[18] Spectral Angle Mapper, [Online]. Available: //step.esa.int/main/wp-content/help/ versions/9.0.0/snap-toolboxes/org.esa.s2tbx. s2tbx.spectral.angle.mapper.ui/sam/SAMProcessor.html
[19] E. Nyamjargal, A. Enkhmanlai, D. Amarsaikhan, and S. Enkhtuya, "Evaluation of principal components for land cover discrimination using object-based classification", Full paper published in eProceedings of the Asian Conference on RS, Ulaanbaatar, Mongolia, 2022.
[20] Y. W. Chang, C. J. Hsieh, K. W. Chang, M. Ringgaard, and C. J. Lin, “Training and testing low-degree polynomial data mappings via linear SVM,” J. Mach. Learn. Res., vol. 11, no. April, pp. 1471–1490, 2010.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Amarsaikhan Damdinsuren, Odontuya Gendaram, Damdinsuren Enkhjargal, Tsogzol Gurjav, Jargaldalai Enkhtuya, Ochirhuyag Lkhamjav
This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright on any research article in the Mongolian Journal of Geography and Geoecology is retained by the author(s).
The authors grant the Mongolian Journal of Geography and Geoecology a license to publish the article and identify itself as the original publisher.
Articles in the Mongolian Journal of Geography and Geoecology are Open Access articles published under a Creative Commons Attribution 4.0 International License CC BY.
This license permits use, distribution and reproduction in any medium, provided the original work is properly cited.
