An object classification technique on aerial and space imagery under low separability of recognition features
DOI:
https://doi.org/10.36023/ujrs.2023.10.3.244Keywords:
aerial and space images, unsupervised classification, supervised classification, clustering, training sample, training sample separabilityAbstract
Classification methods are used in remote sensing of the Earth to study natural resources, monitor the environment, and solve many other problems. Also, they play an important role in involving the remotely sensed data in solving tasks related to achieving sustainable development goals. Among them are supervised and unsupervised classification methods on aerial and space imagery. However, for most thematic problems, it is advisable to use precisely supervised classification methods because the considered problems require setting the characteristics of the initial classes. In supervised classification methods, the features are given by the training sample. Among the well-known approaches to processing the training sample, the following can be distinguished: cluster sampling; approaches to reduce the size of the training sample; approaches that detect representatives who were assigned to the sample of the wrong class to which their class membership corresponds. However, their common disadvantages is that they do not consider the factor of separation of the training sample. This property directly affects the reliability of the classification. The research proposed a technique to increase the reliability of object classification on aerial and space imagery by increasing the separability of the training sample. This technique includes a method of assessing the separability of the training sample. At the same time, it is possible to assess the separability of both two separate classes and the entire set of the educational sample. The developed technique has two branches of application: reducing the size of the training sample and clustering the training sample. In the study, the effectiveness of this technique was experimentally tested on three examples. In two examples, the technique was used to cluster the training sample. In one of these examples, the overall accuracy of the classification increased by 4% (from 77% to 81%), and in the second one by 20% (from 63% to 83%). A reduction in the size of the training sample was applied to the third example. As a result, the dimensionality of the input data was reduced from 167 to 57 layers. That is, the dimensionality decreased by 2.92 times. Also, the overall accuracy of the classification was increased by 2% (from 91% to 93%).
References
Andreiev, A. A. (2020). Hybrid approach to classification of remote sensing data. CERes Journal, 6(2), 32–37.
Andreiev, A., & Kozlova, A. (2021). Enhancement of land cover classification by training samples clustering. In Pattern Recognition and Information Processing (PRIP'2021) : Proceedings of the 15th International Conference, 223-227. Minsk, Belarus: UIIP NASB. ISBN 978-985-7198-07-8.
Andries A, Morse S, Murphy R et al (2019) Translation of Earth observation data into sustainable development indicators: an analytical framework. Sustain Dev, 27, 366–376. https://doi.org/10.1002/sd.1908
Berry, M., Mohamed, A., & Yap, B. W. (2019). Supervised and Unsupervised Learning for Data Science. Springer, Cham.
Bishop, Y. M., Fienberg, S. E., & Holland, P. W. (2007). Discrete multivariate analysis: Theory and Practice. Springer Science & Business Media.
Bruzzone, L. & Demir, B. A. (2014). “A review of modern approaches to classification of remote sensing data”, in Land use and land cover mapping in Europe, I. Manakos, M. Braun, Eds. Springer: Dordrecht, Netherlands, 127–143.
Ferreira, B., Silva, R. G., & Pereira, V. (2017). Feature selection using non-binary decision trees applied to condition monitoring. 22nd IEEE International Conference on Emerging Technologies and Factory Automation (ETFA). https://doi.org/10.1109/etfa.2017.8247642
Green, A., Berman, M., Switzer, P., & Craig, M. (1988). A transformation for ordering multispectral data in terms of image quality with implications for noise removal. IEEE Transactions on Geoscience and Remote Sensing, 26(1), 65–74. https://doi.org/10.1109/36.3001
Huang, H., Shi, G., He, H., Duan, Y., & Luo, F. (2019). Dimensionality reduction of hyperspectral imagery based on spatial–spectral manifold learning. IEEE transactions on cybernetics, 50(6), 2604-2616.
Jain, A. K., & Dubes, R. C. (1988). Algorithms for Clustering Data. Prentice-Hall, Inc., Upper Saddle River, NJ, USA.
Kang, X., Xiang, X., Li, S., & Benediktsson, J. A. (2018). Detection and correction of mislabeled training samples for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing, 56(10), 5673–5686. https://doi.org/10.1109/tgrs.2018.2823866
Landry T, Sotir M, Rajotte J-F, et al (2019) Applying machine learning to earth observations in a standards based workflow. IGARSS 2019–2019 IEEE International Geoscience and Remote Sensing Symposium. IEEE, 5567–5570
Li, H., Cui, J., Zhang, X., Han, Y., & Cao, L. (2022). Dimensionality reduction and classification of hyperspectral remote sensing image feature extraction. Remote Sensing, 14(18), 4579. https://doi.org/10.3390/rs14184579
Lillesand, T. M., Kiefer, R. W., & Chipman, J. W. (2014). Remote Sensing and Image Interpretation. Wiley
Luo, F., Zhang, L., Du, B., & Zhang, L. (2020). Dimensionality reduction with enhanced hybrid-graph discriminant learning for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing, 58(8), 5336-5353.
Melesse, A. M., Weng, Q., Thenkabail, P. S., & Senay, G. B. (2007). Remote sensing sensors and applications in environmental resources mapping and modelling. Sensors, 7(12), 3209–3241. https://doi.org/10.3390/s7123209.
Montero, D., Kraemer, G., Anghelea, A., Camacho, C. A., Brandt, G., Camps-Valls, G., Cremer, F., Flik, I., Gans, F., Habershon, S., Ji, C., Kattenborn, T., Martínez-Ferrer, L., Martinuzzi, F., Reinhardt, M., Söchting, M., Teber, K., & Mahecha, M. (2023a). Data Cubes for Earth System research: Challenges ahead. EarthArXiv (California Digital Library). https://doi.org/10.31223/x58m2v
Omran, M. G. H., Engelbrecht, A. P., & Salman, A. A. (2007). An overview of clustering methods. Intelligent Data Analysis, 11(6), 583–605. https://doi.org/10.3233/ida-2007-11602
Popov, M. O. (2007). Methodology of accuracy assessment of classification of objects on space images. Journal of Automation and Information Sciences, 39, 1–10. https://doi.org/10.1615/J Automat Inf Scien.v39.i1.50
Ruppert, D. (2004). The Elements of Statistical Learning: Data Mining, Inference, and Prediction. Journal of the American Statistical Association, 99(466), 567. https://doi.org/10.1198/jasa.2004.s339
Salem, N., & Hussein, S. (2019). Data dimensional reduction and principal components analysis. Procedia Computer Science, 163, 292–299. https://doi.org/10.1016/j.procs.2019.12.111
Scott, G., & Rajabifard, A. (2017). Sustainable development and geospatial information: a strategic framework for integrating a global policy agenda into national geospatial capabilities. Geo-spatial Information Science, 20(2), 59–76. https://doi.org/10.1080/10095020.2017.1325594
Sedgwick, P. (2014). Cluster sampling. BMJ, 34, 1215. https://doi.org/10.1136/bmj.g1215
Starovoitov, В. В., & Golub, Y. I. (2020). Comparative study of quality estimation of binary classification. Informatika, 17(1), 87–101. https://doi.org/10.37661/1816-0301-2020-17-1-87-101
Subbotin, S. (2010). The training set quality measures for neural network learning. Optical Memory and Neural Networks, 19(2), 126–139. https://doi.org/10.3103/s1060992x10020037
Downloads
Published
Issue
Section
License
Licensing conditions: the authors retain their copyrights and grant the journal the right of first publication of a work, simultaneously licensed in accordance with the Creative Commons Attribution License International CC-BY, which allows you to share the work with proof of authorship of the work and initial publication in this journal.
The authors, directing the manuscript to the editorial office of the Ukrainian Journal of Remote Sensing of the Earth, agree that the editorial board transfers the rights to protection and use of the manuscript (material submitted to the journal editorial board, including such protected copyright objects as photographs of the author, drawings, charts, tables, etc.), including reproduction in print and on the Internet; for distribution; to translate the manuscript into any languages; export and import of copies of the journal with the article of the authors for the purpose of distribution, informing the public. The above rights are transferred by the authors to the editors, without limitation of their validity, and in the territory of all countries of the world without limitation, including in Ukraine.
The authors guarantee that they have exclusive rights to use the submitted material. The editors are not liable to third parties for breach of data by the authors of the guarantees. The authors retain the right to use the published material, its fragments and parts for personal, including scientific and educational purposes. The rights to the manuscript are considered to be transferred by the authors of the editorial board from the moment of the publication of the issue of the journal in which it is published. Reprinting of materials published in the journal by other individuals and legal entities is possible only with the consent of the publisher, with the obligatory indication of the issue of the journal in which the material was published.