Improvement of land cover classification accuracy by training sample clustering

Artem Andreiev, Leonid Artiushyn

Abstract


The subject of this article is land cover classification based on geospatial data. The supervised classification methods are appropriate for most of the thematic tasks of remote sensing because they provide the opportunity to set the characteristics of the initial classes in the form of a training sample set, in contrast to unsupervised methods. There are many approaches to processing such a set; however, their common disadvantage is that they do not consider the factor of training sample separability. This characteristic indicates the extent to which signatures representing different classes do not overlap. A low degree of separability is inherent in high-level training sample mixing. Thus, separability affects classification accuracy. One possible ways to increase separability is training sample clustering. Considering the above, the goal of this study is to develop a training sample clustering technique to improve land cover classification accuracy by increasing the separability of training samples. The tasks of this work are as follows: 1) develop a method for training sample separability assessment; 2) develop a training sample clustering technique based on training sample separability; 3) test the effectiveness of the developed technique by applying it to experimental land cover classification. In the experiments, two land cover classifications were obtained for each of the two selected study areas (i.e., one before and another after training sample clustering. Six land cover classes were defined for each experiment. The training samples were selected for each class. Conclusions. After the application of the developed technique, an increase in the separability of the training samples was evidenced by the developed separability index. In turn, this approach led to an improvement in land cover classification. For the first experiment, this was evidenced by an increase in the overall accuracy and kappa coefficient by 20% (from 63 to 83%) and 21% (from 60% to 81%), respectively. In the second experiment, the increase was 4% (from 77% to 81%) and 5% (from 66% to 71%), respectively.


Keywords


classification; supervised classification; unsupervised classification; clustering; remote sensing; training sample; training sample separability

Full Text:

PDF

References


Melesse, A. M., Weng, Q., Thenkabail, P. S., & Senay, G. B. Remote sensing sensors and applications in environmental resources mapping and modelling. Sensors, 2007, vol. 7, no. 12, pp. 3209-3241. DOI: 10.3390/s7123209.

Zaitseva, E., Stankevich, S., Kozlova, A., Piestova, I., Levashenko, V., & Rusnak, P. Assessment of the risk of disturbance impact on primeval and managed forests based on earth observation data using the example of Slovak Eastern Carpathians. IEEE Access, 2021, vol. 9, pp. 162847-162856. DOI: 10.1109/access.2021.3134375.

Kozlova, A., Stankevich, S., Svideniuk, M., & Andreiev, A. Quantitative Assessment of Forest Dis-turbance with C-Band SAR Data for Decision Making Support in Forest Management. In: Lecture notes on data engineering and communications technologies, 2021, pp 548-562. DOI: 10.1007/978-3-030-82014-5_37.

Stankevich, S. A., & Kozlova, A. A. Long-term land cover change computer-aided mapping by remote sensed imagery, 2015 International Conference on Information and Digital Technologies, Zilina, Slovakia, 2015, pp. 339-341. DOI: 10.1109/DT.2015.7222994.

Kozlova, A., Khyzhniak, A., Piestova, I., & Andreiev, A. Synergetic use of Sentinel-1 and Sentinel-2 data for analysis of urban development and green spaces. 17th International Conference on Geoinformatics - Theoretical and Applied Aspects, 2018, vol. 2018, pp. 1-6. DOI: 10.3997/2214-4609.201801846.

Eddahby, L., Kozlova, A., Popov, M., Lubskiy, N., Mezzane, D., Lukynchuk, I., Larabi, A., & Ibouh, H. Synergetic use of Sentinel-1 and Sentinel-2 data for extraction of built-up area in a rocky desert oasis, example for Draa Tafilalt, south-east of Morocco. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2019, vol. 42, no. 4, pp. 65-68. DOI: 10.5194/isprs-archives-xlii-4-w12-65-2019.

Popov, M. A., Stankevich, S. A., Mosov, S. P., Titarenko, O. V., Dugin, S. S., Golubov, S. I., & Andreiev, A. A. Method for Minefields Mapping by Imagery from Unmanned Aerial Vehicle. Advances in Military Technology, 2022, vol. 17, no. 2, pp. 211-229. DOI: 10.3849/aimt.01722.

Popov, M. A., Тopolnytskyi, М. V., Titarenko, O. V., Stankevich, S. A., & Andreiev, A. A. Forecasting gas and oil potential of subsoil plots via co-analysis of satellite, geological, geophysical and geochemical information by means of subjective logic. WSEAS Transactions on Computer Research, 2020, vol. 8, pp. 90-101. DOI: 10.37394/232018.2020.8.11.

Andries, A., Morse, S., Murphy, R., Lynch, J, Woolliams, E., & Fonweban, J. Translation of Earth observation data into sustainable development indi-cators: An analytical framework. Sustainable Devel-opment, 2018, vol. 27, no. 3, pp. 366-376. DOI: 10.1002/sd.1908.

Scott, G., & Rajabifard, A. Sustainable de-velopment and geospatial information: a strategic framework for integrating a global policy agenda into national geospatial capabilities. Geospatial Information Science, 2017, vol. 20, no. 2, pp. 59-76. DOI: 10.1080/10095020.2017.1325594.

Popov, M., Stankevich, S., Kostyuchenko, Y., & Kozlova, A. Analysis of Local Climate Variations Using Correlation between Satellite Measurements of Methane Emission and Temperature Trends within Physiographic Regions of Ukraine. International Journal of Mathematical, Engineering and Management Sciences, 2019, vol. 4, no. 2, pp. 276-288. DOI: 10.33889/IJMEMS.2019.4.2-023.

Dubovyk, O. The role of Remote Sensing in land degradation assessments: opportunities and challenges. European Journal of Remote Sensing, 2017, vol. 50, no. 1, pp. 601-613. DOI: 10.1080/22797254.2017.1378926.

Popov, M., Kussul, N., Stankevich, S., Kozlova, A., Shelestov, A., Kravchenko, O., Korbakov, M., & Skakun, S. Web service for biodiversity estimation using remote sensing data. International Journal of Digital Earth, 2008, vol. 1, no. 4, pp. 367-376. DOI: 10.1080/17538940802483745.

Eddahby, L., Popov, M., Stankevich, S., Ko-zlova, A., Svideniuk, M., Mezzane, D., Lukyanchuk, I., Larabi, A., & Ibouh, H. Assessing vegetation structural changes in oasis agro-ecosystems using Sentinel-2 image time series: case study for Drâa-tafilalet region Morocco. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2019, vol. 42, no. 4, pp. 69-73. DOI: 10.5194/isprs-archives-xlii-4-w12-69-2019.

Stankevich, S., Piestova, I., Kozlova, A., & Titarenko, O. Satellite Radar Interferometry Processing and Elevation Change Analysis for Geoenvironmental Hazard Assessment. Techniques for Disaster Risk Management and Mitigation, 2020, pp. 127-139. DOI: 10.1002/9781119359203.ch10.

Sathya, R., & Abraham, A. Comparison of supervised and unsupervised learning algorithms for pattern classification. International Journal of Advanced Research in Artificial Intelligence, 2013, vol. 2, no. 2. DOI: 10.14569/ijarai.2013.020206.

Sedgwick, P. Cluster sampling. BMJ, 2014, article no.1215. DOI: 10.1136/bmj.g1215.

Islam, R., Siddiqa, A., Afjal, M. I., Uddin, M. P., & Ulhaq, A. Hyperspectral image classification via information theoretic dimension reduction. Remote Sensing, 2023, vol. 16, no. 5, article no. 807. DOI: 10.3390/rs15041147.

Naumenko, V., Kovalenko, B., & Lukin, V. BPG-based compression analysis of Poissonnoisy medical images. Radioelectronic and Computer Sys-tems, 2023, vol. 3, pp. 91-100. DOI: 10.32620/reks.2023.3.08.

Rebrov, V., & Lukin, V. Post-processing of compressed noisy images by BM3D filter. Radioelec-tronic and Computer Systems, 2023, vol. 4, pp. 100-111. DOI: 10.32620/reks.2023.4.09.

Han, H., Li, D., Liu, W., Zhang, H., & Wang, J. High dimensional mislabeled learning. Neurocom-puting, 2024, vol. 573, article no. 127218. DOI: 10.1016/j.neucom.2023.127218.

Zahorodnia, D., Pigovsky, Y., & Bykovyy, P. Canny-based method of image contour segmentation. International Journal of Computing, 2016, vol. 15, no. 3, pp. 200-205. DOI: 10.47839/ijc.15.3.853.

Marakhimov, A. R., & Khudaybergenov, K. K. Approach to the synthesis of neural network structure during classification. International Journal of Computing, 2020, vol. 19, no, 1, pp. 20-26. DOI: 10.47839/ijc.19.1.1689.

Dehkordi, A. T., Zoej, M. J. V., Ghasemi, H., Ghaderpour, E., & Hassan, Q. K. A New Clustering Method to Generate Training Samples for Supervised Monitoring of Long-Term Water Surface Dynamics Using Landsat Data through Google Earth Engine. Sustainability, 2022, vol. 14, no. 13, article no. 8046. DOI: 10.3390/su14138046.

Clark, A., Phinn, S., & Scarth, P. Pre-Processing training data improves accuracy and generalisability of convolutional neural network based landscape semantic segmentation. Land, 2022, vol. 12, no. 7, article no. 1268. DOI: 10.3390/land12071268.

Andreiev, A., Kozlova, A., Artiushyn, L., & Sedlacek, P. Enhancement of Land Cover Classification by Geospatial Data Cube Optimization. Computer Modeling and Intelligent Systems, 2024, vol. 3702, pp.292-304.

Popov, M. O. Metodologiya otsenki tochnosti klassifikatsii ob"yektov na kosmicheskikh izobrazheniyakh [Methodology of accuracy assessment of classification of objects on space images]. Journal of Automation and Information Sciences, 2007, no. 1, pp. 97-103. Available at: https://www.researchgate.net/publication/245407157_Methodology_of_Accuracy_Assessment_of_Classification_of_Objects_on_Space_Images?enrichId=rgreq-391ade514bc4417ce4facae2236a8bcd-XXX&enrichSource=Y292ZXJQYWdlOzI0NTQwNzE1NztBUzozMDE0NDc5MzA0OTkwNzNAMTQ0ODg4MjE5NTQyMw%3D%3D&el=1_x_2&_esc=publicationCoverPdf. (accessed 12.2.2024) (In Russian)

Vujović, Ž. Classification model evaluation metrics. International Journal of Advanced Computer Science and Applications/International Journal of Advanced Computer Science & Applications, 2021, vol. 12, no. 6. DOI: 10.14569/ijacsa.2021.0120670.

Omran, M. G., Engelbrecht, A. P., & Salman, A. A. An overview of clustering methods. Intelligent Data Analysis, 2007, vol. 11, no. 6, pp. 583-605. DOI: 10.3233/ida-2007-11602.

Shah, S. A., Kiran, M., Nazir, A., & Ashrafani, S. H. Exploring NDVI and NDBI relationship using Landsat 8 OLI/TIRS in Khangarh Taluka, Ghotki. Malaysian Journal of Geosciences, 2022, vol. 6, no. 1, pp. 8-11. DOI: 10.26480/mjg.01.2022.08.11.

De Souza, C., Mercante, E., Prudente, V., & Justina, D. Methods of performance evaluation for the supervised classification of satellite imagery in determining land cover classes. Cienc. Inv. Agr., 2013, vol. 40, no. 2, pp. 419-428. DOI: 10.4067/s0718-16202013000200016.




DOI: https://doi.org/10.32620/reks.2024.2.06

Refbacks

  • There are currently no refbacks.