|
Registro Completo |
|
Biblioteca(s): |
Embrapa Amazônia Oriental. |
|
Data corrente: |
14/07/2022 |
|
Data da última atualização: |
14/07/2022 |
|
Tipo da produção científica: |
Artigo em Periódico Indexado |
|
Autoria: |
BATISTA, J. E.; RODRIGUES, N. M.; CABRAL, A. I. R.; VASCONCELOS, M. J. P.; VENTURIERI, A.; SILVA, L. G. T.; SILVA, S. |
|
Afiliação: |
JOÃO E. BATISTA, University of Lisbon; NUNO M. RODRIGUES, University of Lisbon; ANA I. R. CABRAL, University of Lisbon; MARIA J. P. VASCONCELOS, University of Lisbon; ADRIANO VENTURIERI, CPATU; LUIZ GUILHERME TEIXEIRA SILVA, CPATU; SARA SILVA, University of Lisbon. |
|
Título: |
Optical time series for the separation of land cover types with similar spectral signatures: cocoa agroforest and forest. |
|
Ano de publicação: |
2022 |
|
Fonte/Imprenta: |
International Journal of Remote Sensing, v. 43, n. 9, p. 3298-3319, 2022. |
|
DOI: |
https://doi.org/10.1080/01431161.2022.2089540 |
|
Idioma: |
Inglês |
|
Conteúdo: |
One of the main applications of machine learning (ML) in remote sensing (RS) is the pixel-level classification of satellite images into land cover types. Although classes with different spectral signatures can be easily separated, e.g. aquatic and terrestrial land cover types, others have similar spectral signatures and are hard to separate using only the information within a single pixel. This work focused on the separation of two cover types with similar spectral signatures, cocoa agroforest and forest, over an area in Pará, Brazil. For this, we study the training and application of several ML algorithms on datasets obtained from a single composite image, a time-series (TS) composite obtained from the same location and by preprocessing the TS composite using simple TS preprocessing techniques. As expected, when ML algorithms are applied to a dataset obtained from a composite image, the median producer's accuracy (PA) and user's accuracy (UA) in those two classes are significantly lower than the median overall accuracy (OA) for all classes. The second dataset allows the ML models to learn the evolution of the spectral signatures over 5 months. Compared to the first dataset, the results indicate that ML models generalize better using TS data, even if the series are short and without any preprocessing. This generalization is further improved in the last dataset. The ML models are subsequently applied to an area with different geographical bounds. These last results indicate that, out of seven classifiers, the popular random forest (RF) algorithm ranked fourth, while XGBoost (×GB) obtained the best results. The best OA, as well as the best PA/UA balance, were obtained by performing feature construction using the M3GP algorithm and then applying XGB to the new extended dataset. MenosOne of the main applications of machine learning (ML) in remote sensing (RS) is the pixel-level classification of satellite images into land cover types. Although classes with different spectral signatures can be easily separated, e.g. aquatic and terrestrial land cover types, others have similar spectral signatures and are hard to separate using only the information within a single pixel. This work focused on the separation of two cover types with similar spectral signatures, cocoa agroforest and forest, over an area in Pará, Brazil. For this, we study the training and application of several ML algorithms on datasets obtained from a single composite image, a time-series (TS) composite obtained from the same location and by preprocessing the TS composite using simple TS preprocessing techniques. As expected, when ML algorithms are applied to a dataset obtained from a composite image, the median producer's accuracy (PA) and user's accuracy (UA) in those two classes are significantly lower than the median overall accuracy (OA) for all classes. The second dataset allows the ML models to learn the evolution of the spectral signatures over 5 months. Compared to the first dataset, the results indicate that ML models generalize better using TS data, even if the series are short and without any preprocessing. This generalization is further improved in the last dataset. The ML models are subsequently applied to an area with different geographical bounds. These last results indicate t... Mostrar Tudo |
|
Palavras-Chave: |
Áreas tropicais; Séries temporais. |
|
Thesagro: |
Cacau; Cobertura do Solo. |
|
Categoria do assunto: |
-- |
|
Marc: |
LEADER 02632naa a2200253 a 4500
001 2144723
005 2022-07-14
008 2022 bl uuuu u00u1 u #d
024 7 $ahttps://doi.org/10.1080/01431161.2022.2089540$2DOI
100 1 $aBATISTA, J. E.
245 $aOptical time series for the separation of land cover types with similar spectral signatures$bcocoa agroforest and forest.$h[electronic resource]
260 $c2022
520 $aOne of the main applications of machine learning (ML) in remote sensing (RS) is the pixel-level classification of satellite images into land cover types. Although classes with different spectral signatures can be easily separated, e.g. aquatic and terrestrial land cover types, others have similar spectral signatures and are hard to separate using only the information within a single pixel. This work focused on the separation of two cover types with similar spectral signatures, cocoa agroforest and forest, over an area in Pará, Brazil. For this, we study the training and application of several ML algorithms on datasets obtained from a single composite image, a time-series (TS) composite obtained from the same location and by preprocessing the TS composite using simple TS preprocessing techniques. As expected, when ML algorithms are applied to a dataset obtained from a composite image, the median producer's accuracy (PA) and user's accuracy (UA) in those two classes are significantly lower than the median overall accuracy (OA) for all classes. The second dataset allows the ML models to learn the evolution of the spectral signatures over 5 months. Compared to the first dataset, the results indicate that ML models generalize better using TS data, even if the series are short and without any preprocessing. This generalization is further improved in the last dataset. The ML models are subsequently applied to an area with different geographical bounds. These last results indicate that, out of seven classifiers, the popular random forest (RF) algorithm ranked fourth, while XGBoost (×GB) obtained the best results. The best OA, as well as the best PA/UA balance, were obtained by performing feature construction using the M3GP algorithm and then applying XGB to the new extended dataset.
650 $aCacau
650 $aCobertura do Solo
653 $aÁreas tropicais
653 $aSéries temporais
700 1 $aRODRIGUES, N. M.
700 1 $aCABRAL, A. I. R.
700 1 $aVASCONCELOS, M. J. P.
700 1 $aVENTURIERI, A.
700 1 $aSILVA, L. G. T.
700 1 $aSILVA, S.
773 $tInternational Journal of Remote Sensing$gv. 43, n. 9, p. 3298-3319, 2022.
Download
Esconder MarcMostrar Marc Completo |
|
Registro original: |
Embrapa Amazônia Oriental (CPATU) |
|
