Tree Species Classification of Drone Hyperspectral and RGB Imagery with Deep Learning Convolutional Neural Networks
Interest in drone solutions in forestry applications is growing. Using drones, datasets can be captured flexibly and at high spatial and temporal resolutions when needed. In forestry applications, fundamental tasks include the detection of individual trees, tree species classification, biomass estim...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-03-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/12/7/1070 |
| _version_ | 1798025831866433536 |
|---|---|
| author | Somayeh Nezami Ehsan Khoramshahi Olli Nevalainen Ilkka Pölönen Eija Honkavaara |
| author_facet | Somayeh Nezami Ehsan Khoramshahi Olli Nevalainen Ilkka Pölönen Eija Honkavaara |
| author_sort | Somayeh Nezami |
| collection | DOAJ |
| description | Interest in drone solutions in forestry applications is growing. Using drones, datasets can be captured flexibly and at high spatial and temporal resolutions when needed. In forestry applications, fundamental tasks include the detection of individual trees, tree species classification, biomass estimation, etc. Deep neural networks (DNN) have shown superior results when comparing with conventional machine learning methods such as multi-layer perceptron (MLP) in cases of huge input data. The objective of this research is to investigate 3D convolutional neural networks (3D-CNN) to classify three major tree species in a boreal forest: pine, spruce, and birch. The proposed 3D-CNN models were employed to classify tree species in a test site in Finland. The classifiers were trained with a dataset of 3039 manually labelled trees. Then the accuracies were assessed by employing independent datasets of 803 records. To find the most efficient set of feature combination, we compare the performances of 3D-CNN models trained with hyperspectral (HS) channels, Red-Green-Blue (RGB) channels, and canopy height model (CHM), separately and combined. It is demonstrated that the proposed 3D-CNN model with RGB and HS layers produces the highest classification accuracy. The producer accuracy of the best 3D-CNN classifier on the test dataset were 99.6%, 94.8%, and 97.4% for pines, spruces, and birches, respectively. The best 3D-CNN classifier produced ~5% better classification accuracy than the MLP with all layers. Our results suggest that the proposed method provides excellent classification results with acceptable performance metrics for HS datasets. Our results show that pine class was detectable in most layers. Spruce was most detectable in RGB data, while birch was most detectable in the HS layers. Furthermore, the RGB datasets provide acceptable results for many low-accuracy applications. |
| first_indexed | 2024-04-11T18:25:15Z |
| format | Article |
| id | doaj.art-57a86e8177244eee85c890d8041f57ab |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-04-11T18:25:15Z |
| publishDate | 2020-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-57a86e8177244eee85c890d8041f57ab2022-12-22T04:09:38ZengMDPI AGRemote Sensing2072-42922020-03-01127107010.3390/rs12071070rs12071070Tree Species Classification of Drone Hyperspectral and RGB Imagery with Deep Learning Convolutional Neural NetworksSomayeh Nezami0Ehsan Khoramshahi1Olli Nevalainen2Ilkka Pölönen3Eija Honkavaara4Department of Remote Sensing and Photogrammetry of the Finnish Geospatial Research Institute FGI, Geodeetinrinne 2, FI-02430 Masala, FinlandDepartment of Remote Sensing and Photogrammetry of the Finnish Geospatial Research Institute FGI, Geodeetinrinne 2, FI-02430 Masala, FinlandFinnish Meteorological Institute, Climate System Research, 00100 Helsinki, FinlandFaculty of Information Technology, University of Jyväskylä, Seminaarinkatu 15, 40014 Jyväskylän yliopisto, FinlandDepartment of Remote Sensing and Photogrammetry of the Finnish Geospatial Research Institute FGI, Geodeetinrinne 2, FI-02430 Masala, FinlandInterest in drone solutions in forestry applications is growing. Using drones, datasets can be captured flexibly and at high spatial and temporal resolutions when needed. In forestry applications, fundamental tasks include the detection of individual trees, tree species classification, biomass estimation, etc. Deep neural networks (DNN) have shown superior results when comparing with conventional machine learning methods such as multi-layer perceptron (MLP) in cases of huge input data. The objective of this research is to investigate 3D convolutional neural networks (3D-CNN) to classify three major tree species in a boreal forest: pine, spruce, and birch. The proposed 3D-CNN models were employed to classify tree species in a test site in Finland. The classifiers were trained with a dataset of 3039 manually labelled trees. Then the accuracies were assessed by employing independent datasets of 803 records. To find the most efficient set of feature combination, we compare the performances of 3D-CNN models trained with hyperspectral (HS) channels, Red-Green-Blue (RGB) channels, and canopy height model (CHM), separately and combined. It is demonstrated that the proposed 3D-CNN model with RGB and HS layers produces the highest classification accuracy. The producer accuracy of the best 3D-CNN classifier on the test dataset were 99.6%, 94.8%, and 97.4% for pines, spruces, and birches, respectively. The best 3D-CNN classifier produced ~5% better classification accuracy than the MLP with all layers. Our results suggest that the proposed method provides excellent classification results with acceptable performance metrics for HS datasets. Our results show that pine class was detectable in most layers. Spruce was most detectable in RGB data, while birch was most detectable in the HS layers. Furthermore, the RGB datasets provide acceptable results for many low-accuracy applications.https://www.mdpi.com/2072-4292/12/7/1070deep learningdrone imageryhyperspectral image classificationtree species classification3d convolutional neural networks |
| spellingShingle | Somayeh Nezami Ehsan Khoramshahi Olli Nevalainen Ilkka Pölönen Eija Honkavaara Tree Species Classification of Drone Hyperspectral and RGB Imagery with Deep Learning Convolutional Neural Networks Remote Sensing deep learning drone imagery hyperspectral image classification tree species classification 3d convolutional neural networks |
| title | Tree Species Classification of Drone Hyperspectral and RGB Imagery with Deep Learning Convolutional Neural Networks |
| title_full | Tree Species Classification of Drone Hyperspectral and RGB Imagery with Deep Learning Convolutional Neural Networks |
| title_fullStr | Tree Species Classification of Drone Hyperspectral and RGB Imagery with Deep Learning Convolutional Neural Networks |
| title_full_unstemmed | Tree Species Classification of Drone Hyperspectral and RGB Imagery with Deep Learning Convolutional Neural Networks |
| title_short | Tree Species Classification of Drone Hyperspectral and RGB Imagery with Deep Learning Convolutional Neural Networks |
| title_sort | tree species classification of drone hyperspectral and rgb imagery with deep learning convolutional neural networks |
| topic | deep learning drone imagery hyperspectral image classification tree species classification 3d convolutional neural networks |
| url | https://www.mdpi.com/2072-4292/12/7/1070 |
| work_keys_str_mv | AT somayehnezami treespeciesclassificationofdronehyperspectralandrgbimagerywithdeeplearningconvolutionalneuralnetworks AT ehsankhoramshahi treespeciesclassificationofdronehyperspectralandrgbimagerywithdeeplearningconvolutionalneuralnetworks AT ollinevalainen treespeciesclassificationofdronehyperspectralandrgbimagerywithdeeplearningconvolutionalneuralnetworks AT ilkkapolonen treespeciesclassificationofdronehyperspectralandrgbimagerywithdeeplearningconvolutionalneuralnetworks AT eijahonkavaara treespeciesclassificationofdronehyperspectralandrgbimagerywithdeeplearningconvolutionalneuralnetworks |