Assessing the Performance of Machine Learning Algorithms for Urban Land Cover Classification using Multispectral Satellite Imagery

M. B Poku, Yakubu Issaka, Y. Y. Ziggah

Abstract


Land use and land cover mapping is a critical process in effective land management, providing valuable insights into the spatial distribution and characteristics of different land uses and covers within a region. With the advancements in geospatial technology and the accessibility to high-resolution satellite imagery, various classification algorithms have emerged as powerful tools for mapping and analysing land cover patterns. The selection of a specific classification algorithm significantly influences the accuracy and reliability of the obtained results, thereby impacting the effectiveness of decision-making based on the classification outcomes. Aside the traditional classification techniques such as maximum likelihood, minimum distance and the parallelepiped classification algorithms, various machine learning methods have emerged for image classification. Machine learning techniques offer valuable advantages due to their capacity to learn from data, adapt to new datasets, and achieve good generalisation performance. This paper conducted a comparative study of four classification algorithms: Support Vector Machine (SVM), Random Forest (RF), K- Nearest Neighbour (KNN) and the Maximum Likelihood Classifier (MLC). A comprehensive dataset comprising of a high-resolution multispectral satellite imagery and ground truth data was employed. The study area is a representative of diverse land cover types including settlement, vegetation, forested, water and bare lands. The accuracy metrics obtained showed that the SVM obtained the best classification performance achieving a precision of 0.84, a recall of 0.82, an F1-Score of 0.83 and an overall accuracy of 0.8932.


Keywords


Multispectral, Machine Learning, Support Vector Machine, K-Nearest Neighbour, Random Forest, Image Classification

Full Text:

PDF

References


Alkaradaghi, K., Ali, S. S., Al-Ansari, N. and Laue, J. (2019), “Land Use Classification and Change Detection Using Multi-temporal Landsat Imagery in Sulaimaniyah Governorate, Iraq”, Advances in Remote Sensing and Geo Informatics Applications, Vol 15, pp 117-120.

Alshari, E. A. and Gawali, B. W. (2021), “Develop-ment of Classification System for LULC using Remote Sensing and GIS”, Global Transitions Proceeding, Vol, pp 8-27.

Amini, S., Saber, M., Rabiei-Dastjerdi, H. and Homayouni, S. (2022). “Urban Land Use and Land Cover Change Analysis Using Random Forest Classification of Landsat Time Series”, Remote Sensing, Vol 14, No. 2654, pp 1-15.

Atef, I., Ahmed, W. and Abdel-Maguid, R. H. (2023), “Modelling of Land Use Land Cover Changes using Machine Learning and GIS Techniques: A Case Study in El Fayoum Gove-rnorate, Egypt”, Environmental Monitoring As-sessment, Vol 195, No. 637, pp 1-15.

Bayes, T. (1764) "An Essay Toward Solving a Problem in the Doctrine of Chances", Philosop-hical Transactions of the Royal Society of London, Vol 53, pp. 370-418.

Breiman, L. (2001), “Random Forests”, Machine Learning, Vol 45, No. 1, pp 5–32.

Deilmei, B. R., Ahmad, B. B. and Zabili, H. (2014), “Comparison of Two Classification Methods MLC and SVM) to Extract Land Use Land Cov-er in Johor, Malaysia”, 7th IGRSM International Remote Sensing and GIS Conference and Exhi-bition, Vol. 20.

Dietterich, T.G. (2000), “An Experimental Comparison of Three Methods for Constructing Ensembles of Decision Trees: Bagging, Boo-sting, and Randomization”, Machine Learning, Vol 40, No.2, pp. 39–157.

Medina, J. A. V. and Atehortua, B. E. A (2019), “Comparison of maximum likelihood, support vector machines, and random forest techniques in satellite images classification”, Tecnura, Vol. 23, No. 59, pp. 13-26.

Noi, P. T. and Kappas, M. (2021), “Comparison of Random Forest, K-Nearest Neighbor, and Sup-port Vector Machine Classifiers for Land Cover Classification Using Sentinel-2 Imagery”, Sen-sors, Vol 18, No. 18, pp 1-20.

Prama, A. A. and Danoedoro, P. (2016), Compa-rison Effectiveness of Pixel Based Classific-ation and Object Based Classification Using High Resolution Image In Floristic Composi-tion Mapping (Study Case: Gunung Tidar Magelang City)”, IOP Conference Series: Earth Enviroment Science, Vol. 47, pp 1-10.

Rodriguez-Galiano, V. F., Ghimire, B., Rogan, J., Chica-Olmo, M. and Rigol,-Sanchez, J.P. (201-2), “An Assessment of the Effectiveness of a Random Forest Classifier for Land-Cover Clas-sification”, ISPRS Journal of Photogra-mmetry and Remote Sensing, Vol. 67, pp. 93-105.

Talukdar S., Singha P., Mahato S., Shahfahad, Pal S., Liou Y., and Rahman A. (2020), “Land-Use Land-Cover Classification by Machine Learn-ing Classifiers for Satellite Observations—A Review, Remote Sens, Vol. 12, No., pp. 1–24.

Tassi, A.; Gigante, D.; Modica, G.; Di Martino, L.; Vizzari, M. (2021), Pixel- vs. Object-Based Landsat 8 Data Classification in Google Earth Engine Using Random Forest: The Case Study of Maiella National Park, Remote Sens., Vol. 13, No.13, pp 1–20.

Ye, S., Pontius, R.G. and Rakshit, R. (2018), “A Review of Accuracy Assessment for Object-Based Image Analysis: from Per-pixel to Per-Polygon Approaches” ISPRS Journal of Phot-ogrammetry and Remote Sensing, Vol. 141, pp. 137-147.

Yuh, Y. Y., Tracz, W., Matthew, H. D. and Turner, S.E. (2022), “Application of Machine Learning Approaches for Land Cover Monitoring in Northern Cameroon”, Ecological Informatics, Vol 72, pp 1-15.


Refbacks

  • There are currently no refbacks.