Tree Species Classification at the Pixel Level Using Deep Learning and Multispectral Time Series in an Imbalanced Context
This paper investigates tree species classification using the Sentinel-2 multispectral satellite image time series (SITS). Despite its importance for many applications and users, such mapping is often unavailable or outdated. The value of using SITS to classify tree species on a large scale has been...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-03-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/17/7/1190 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850212494378795008 |
|---|---|
| author | Florian Mouret David Morin Milena Planells Cécile Vincent-Barbaroux |
| author_facet | Florian Mouret David Morin Milena Planells Cécile Vincent-Barbaroux |
| author_sort | Florian Mouret |
| collection | DOAJ |
| description | This paper investigates tree species classification using the Sentinel-2 multispectral satellite image time series (SITS). Despite its importance for many applications and users, such mapping is often unavailable or outdated. The value of using SITS to classify tree species on a large scale has been demonstrated in numerous studies. However, many methods proposed in the literature still rely on a standard machine learning algorithm, usually the random forest (RF) algorithm. Our analysis shows that the use of deep learning (DL) models can lead to a significant improvement in classification results, especially in an imbalanced context where the RF algorithm tends to predict the majority class. In our case study in central France with 10 tree species, we obtained an overall accuracy (OA) of around 95% and an F1-macro score of around 80% using three different benchmark DL architectures (fully connected, convolutional, and attention-based networks). In contrast, using the RF algorithm, the OA and F1 scores obtained were 92% and 60%, indicating that the minority classes are poorly classified. Our results also show that DL models are robust to imbalanced data, although small improvements can be obtained by specifically addressing this issue. Validation on independent in situ data shows that all models struggle to predict in areas not well covered by training data, but even in this situation, the RF algorithm is largely outperformed by deep learning models for minority classes. The proposed framework can be easily implemented as a strong baseline, even with a limited amount of reference data. |
| format | Article |
| id | doaj-art-e192c614f3d94f6dbe9658a983cef8fc |
| institution | OA Journals |
| issn | 2072-4292 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj-art-e192c614f3d94f6dbe9658a983cef8fc2025-08-20T02:09:20ZengMDPI AGRemote Sensing2072-42922025-03-01177119010.3390/rs17071190Tree Species Classification at the Pixel Level Using Deep Learning and Multispectral Time Series in an Imbalanced ContextFlorian Mouret0David Morin1Milena Planells2Cécile Vincent-Barbaroux3P2E Laboratory (Physiology, Ecology and Environment), USC INRAE 1328, Université of Orléans, 45100 Orléans, FranceCESBIO (Centre d’Etude Spatiale de La Biosphère), CNES/CNRS/INRAE/IRD, Université de Toulouse, 31400 Toulouse, FranceCESBIO (Centre d’Etude Spatiale de La Biosphère), CNES/CNRS/INRAE/IRD, Université de Toulouse, 31400 Toulouse, FranceP2E Laboratory (Physiology, Ecology and Environment), USC INRAE 1328, Université of Orléans, 45100 Orléans, FranceThis paper investigates tree species classification using the Sentinel-2 multispectral satellite image time series (SITS). Despite its importance for many applications and users, such mapping is often unavailable or outdated. The value of using SITS to classify tree species on a large scale has been demonstrated in numerous studies. However, many methods proposed in the literature still rely on a standard machine learning algorithm, usually the random forest (RF) algorithm. Our analysis shows that the use of deep learning (DL) models can lead to a significant improvement in classification results, especially in an imbalanced context where the RF algorithm tends to predict the majority class. In our case study in central France with 10 tree species, we obtained an overall accuracy (OA) of around 95% and an F1-macro score of around 80% using three different benchmark DL architectures (fully connected, convolutional, and attention-based networks). In contrast, using the RF algorithm, the OA and F1 scores obtained were 92% and 60%, indicating that the minority classes are poorly classified. Our results also show that DL models are robust to imbalanced data, although small improvements can be obtained by specifically addressing this issue. Validation on independent in situ data shows that all models struggle to predict in areas not well covered by training data, but even in this situation, the RF algorithm is largely outperformed by deep learning models for minority classes. The proposed framework can be easily implemented as a strong baseline, even with a limited amount of reference data.https://www.mdpi.com/2072-4292/17/7/1190forest monitoringtree species classificationdeep learningmultispectral satellite time seriesimbalanced dataSentinel-2 |
| spellingShingle | Florian Mouret David Morin Milena Planells Cécile Vincent-Barbaroux Tree Species Classification at the Pixel Level Using Deep Learning and Multispectral Time Series in an Imbalanced Context Remote Sensing forest monitoring tree species classification deep learning multispectral satellite time series imbalanced data Sentinel-2 |
| title | Tree Species Classification at the Pixel Level Using Deep Learning and Multispectral Time Series in an Imbalanced Context |
| title_full | Tree Species Classification at the Pixel Level Using Deep Learning and Multispectral Time Series in an Imbalanced Context |
| title_fullStr | Tree Species Classification at the Pixel Level Using Deep Learning and Multispectral Time Series in an Imbalanced Context |
| title_full_unstemmed | Tree Species Classification at the Pixel Level Using Deep Learning and Multispectral Time Series in an Imbalanced Context |
| title_short | Tree Species Classification at the Pixel Level Using Deep Learning and Multispectral Time Series in an Imbalanced Context |
| title_sort | tree species classification at the pixel level using deep learning and multispectral time series in an imbalanced context |
| topic | forest monitoring tree species classification deep learning multispectral satellite time series imbalanced data Sentinel-2 |
| url | https://www.mdpi.com/2072-4292/17/7/1190 |
| work_keys_str_mv | AT florianmouret treespeciesclassificationatthepixellevelusingdeeplearningandmultispectraltimeseriesinanimbalancedcontext AT davidmorin treespeciesclassificationatthepixellevelusingdeeplearningandmultispectraltimeseriesinanimbalancedcontext AT milenaplanells treespeciesclassificationatthepixellevelusingdeeplearningandmultispectraltimeseriesinanimbalancedcontext AT cecilevincentbarbaroux treespeciesclassificationatthepixellevelusingdeeplearningandmultispectraltimeseriesinanimbalancedcontext |