Machine Learning Classification Ensemble of Multitemporal Sentinel-2 Images: The Case of a Mixed Mediterranean Ecosystem

被引:28
|
作者
Vasilakos, Christos [1 ]
Kavroudakis, Dimitris [1 ]
Georganta, Aikaterini [1 ]
机构
[1] Univ Aegean, Dept Geog, Mitilini 81100, Greece
关键词
remote sensing; classification ensemble; machine learning; Sentinel-2; geographic information system (GIS); LAND-COVER CLASSIFICATION; SUPPORT VECTOR MACHINES; ARTIFICIAL NEURAL-NETWORKS; HYPERSPECTRAL IMAGES; RANDOM FOREST; ALGORITHM; SYSTEM; DISCRIMINATION; CLASSIFIERS; ACCURACY;
D O I
10.3390/rs12122005
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Land cover type classification still remains an active research topic while new sensors and methods become available. Applications such as environmental monitoring, natural resource management, and change detection require more accurate, detailed, and constantly updated land-cover type mapping. These needs are fulfilled by newer sensors with high spatial and spectral resolution along with modern data processing algorithms. Sentinel-2 sensor provides data with high spatial, spectral, and temporal resolution for the in classification of highly fragmented landscape. This study applies six traditional data classifiers and nine ensemble methods on multitemporal Sentinel-2 image datasets for identifying land cover types in the heterogeneous Mediterranean landscape of Lesvos Island, Greece. Support vector machine, random forest, artificial neural network, decision tree, linear discriminant analysis, and k-nearest neighbor classifiers are applied and compared with nine ensemble classifiers on the basis of different voting methods. kappa statistic, F1-score, and Matthews correlation coefficient metrics were used in the assembly of the voting methods. Support vector machine outperformed the base classifiers with kappa of 0.91. Support vector machine also outperformed the ensemble classifiers in an unseen dataset. Five voting methods performed better than the rest of the classifiers. A diversity study based on four different metrics revealed that an ensemble can be avoided if a base classifier shows an identifiable superiority. Therefore, ensemble approaches should include a careful selection of base-classifiers based on a diversity analysis.
引用
收藏
页数:25
相关论文
共 50 条
  • [1] Burned Area Classification Based on Extreme Learning Machine and Sentinel-2 Images
    Gajardo, John
    Mora, Marco
    Valdes-Nicolao, Guillermo
    Carrasco-Benavides, Marcos
    [J]. APPLIED SCIENCES-BASEL, 2022, 12 (01):
  • [2] Turbidity classification of the Paraopeba River using machine learning and Sentinel-2 images
    Batista, Leonardo Vidal
    [J]. IEEE LATIN AMERICA TRANSACTIONS, 2022, 20 (05) : 799 - 805
  • [3] BENCHMARK OF MACHINE LEARNING METHODS FOR CLASSIFICATION OF A SENTINEL-2 IMAGE
    Pirotti, F.
    Sunar, F.
    Piragnolo, M.
    [J]. XXIII ISPRS CONGRESS, COMMISSION VII, 2016, 41 (B7): : 335 - 340
  • [4] Automated Machine Learning Driven Stacked Ensemble Modeling for Forest Aboveground Biomass Prediction Using Multitemporal Sentinel-2 Data
    Naik, Parth
    Dalponte, Michele
    Bruzzone, Lorenzo
    [J]. IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2023, 16 : 3442 - 3454
  • [5] Extracting Tea Plantations from Multitemporal Sentinel-2 Images Based on Deep Learning Networks
    Yao, Zhongxi
    Zhu, Xiaochen
    Zeng, Yan
    Qiu, Xinfa
    [J]. AGRICULTURE-BASEL, 2023, 13 (01):
  • [6] Machine Learning for Cloud Detection of Globally Distributed Sentinel-2 Images
    Cilli, Roberto
    Monaco, Alfonso
    Amoroso, Nicola
    Tateo, Andrea
    Tangaro, Sabina
    Bellotti, Roberto
    [J]. REMOTE SENSING, 2020, 12 (15)
  • [7] Machine Learning-Based Classification of Small-Sized Wetlands Using Sentinel-2 Images
    Salas, Eric Ariel L.
    Kumaran, Sakthi Subburayalu
    Bennett, Robert
    Willis, Leeoria P.
    Mitchell, Kayla
    [J]. AIMS GEOSCIENCES, 2024, 10 (01): : 62 - 79
  • [8] Land Cover and Crop Classification using Multitemporal Sentinel-2 Images Based on Crops Phenological Cycle
    Khaliq, Aleem
    Peroni, Leonardo
    Chiaberge, Marcello
    [J]. 2018 IEEE WORKSHOP ON ENVIRONMENTAL, ENERGY, AND STRUCTURAL MONITORING SYSTEMS (EESMS), 2018, : 19 - 23
  • [9] MULTITEMPORAL SEGMENTATION OF SENTINEL-2 IMAGES IN AN AGRICULTURAL INTENSIFICATION REGION IN BRAZIL
    Dos Santos, L. T.
    Werner, J. P. S.
    Dos Reis, A. A.
    Toro, A. P. G.
    Antunes, J. F. G.
    Coutinho, A. C.
    Lamparelli, R. A. C.
    Magalhaes, P. S. G.
    Esquerdo, J. C. D. M.
    Figueiredo, G. K. D. A.
    [J]. XXIV ISPRS CONGRESS: IMAGING TODAY, FORESEEING TOMORROW, COMMISSION III, 2022, 5-3 : 389 - 395
  • [10] Feature Optimization-Based Machine Learning Approach for Czech Land Cover Classification Using Sentinel-2 Images
    Wang, Chunling
    Hang, Tianyi
    Zhu, Changke
    Zhang, Qi
    [J]. APPLIED SCIENCES-BASEL, 2024, 14 (06):