Can we use deep learning models to identify the functionality of plastics from space?
The function of plastics is an important issue, especially since it determines whether or not they can be recycled. This study presents a two-stage workflow to identify the functions of plastic materials on land surfaces using a deep learning model trained with Sentinel-2 satellite images. First, a...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-09-01
|
| Series: | International Journal of Applied Earth Observations and Geoinformation |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1569843223003151 |
| _version_ | 1797677698023161856 |
|---|---|
| author | Shanyu Zhou Lichao Mou, Dr. Yuansheng Hua Lixian Zhang Hermann Kaufmann Xiao Xiang Zhu |
| author_facet | Shanyu Zhou Lichao Mou, Dr. Yuansheng Hua Lixian Zhang Hermann Kaufmann Xiao Xiang Zhu |
| author_sort | Shanyu Zhou |
| collection | DOAJ |
| description | The function of plastics is an important issue, especially since it determines whether or not they can be recycled. This study presents a two-stage workflow to identify the functions of plastic materials on land surfaces using a deep learning model trained with Sentinel-2 satellite images. First, a classification map identifying 10 distinct plastic types was obtained by evaluating spaceborne hyperspectral PRISMA data. Then, different deep learning algorithms were used to assign functions to the initially classified plastic targets based on the RGB information extracted from Sentinel-2 satellite images. A total of 1,645 plastic polygons were manually labeled on RGB images of Sentinel-2, and the following five main function types were identified: plastic cover sheeting for construction areas, greenhouse structures, photovoltaic panels (PVs), roof materials, and sport field floorings. By comparing three state-of-the-art deep learning models, including GoogLeNet, VGGNet, and ResNet, an overall accuracy of 78% was achieved on the test dataset using the VGG-13 network. The model performed well in identifying PVs, greenhouses, and construction sites, with F1 scores of 0.85, 0.77, and 0.71 respectively. The performance of the model in identifying roofs and sport field floorings was lower, with respective F1 scores of 0.57 and 0.59. Overall, the results show that the proposed workflow using deep learning algorithms trained on Sentinel-2 images has a great potential to identify the function of plastic materials on land surfaces. |
| first_indexed | 2024-03-11T22:48:59Z |
| format | Article |
| id | doaj.art-a92eb6a257e34463b5fdccebe3b9e502 |
| institution | Directory Open Access Journal |
| issn | 1569-8432 |
| language | English |
| last_indexed | 2024-03-11T22:48:59Z |
| publishDate | 2023-09-01 |
| publisher | Elsevier |
| record_format | Article |
| series | International Journal of Applied Earth Observations and Geoinformation |
| spelling | doaj.art-a92eb6a257e34463b5fdccebe3b9e5022023-09-22T04:38:22ZengElsevierInternational Journal of Applied Earth Observations and Geoinformation1569-84322023-09-01123103491Can we use deep learning models to identify the functionality of plastics from space?Shanyu Zhou0Lichao Mou, Dr.1Yuansheng Hua2Lixian Zhang3Hermann Kaufmann4Xiao Xiang Zhu5Data Science in Earth Observation, Technical University of Munich, 80333 Munich, Germany; German Research Centre for Geosciences (GFZ), Remote Sensing and Geoinformatics Section, 14473 Potsdam, GermanyData Science in Earth Observation, Technical University of Munich, 80333 Munich, Germany; Corresponding authors.Data Science in Earth Observation, Technical University of Munich, 80333 Munich, Germany; College of Civil and Transportation Engineering, Shenzhen University, 518060 Shenzhen, ChinaData Science in Earth Observation, Technical University of Munich, 80333 Munich, Germany; Ministry of Education Key Laboratory for Earth System Modeling, Department of Earth System Science, Tsinghua University, 100084 Beijing, ChinaGerman Research Centre for Geosciences (GFZ), Remote Sensing and Geoinformatics Section, 14473 Potsdam, GermanyData Science in Earth Observation, Technical University of Munich, 80333 Munich, Germany; Corresponding authors.The function of plastics is an important issue, especially since it determines whether or not they can be recycled. This study presents a two-stage workflow to identify the functions of plastic materials on land surfaces using a deep learning model trained with Sentinel-2 satellite images. First, a classification map identifying 10 distinct plastic types was obtained by evaluating spaceborne hyperspectral PRISMA data. Then, different deep learning algorithms were used to assign functions to the initially classified plastic targets based on the RGB information extracted from Sentinel-2 satellite images. A total of 1,645 plastic polygons were manually labeled on RGB images of Sentinel-2, and the following five main function types were identified: plastic cover sheeting for construction areas, greenhouse structures, photovoltaic panels (PVs), roof materials, and sport field floorings. By comparing three state-of-the-art deep learning models, including GoogLeNet, VGGNet, and ResNet, an overall accuracy of 78% was achieved on the test dataset using the VGG-13 network. The model performed well in identifying PVs, greenhouses, and construction sites, with F1 scores of 0.85, 0.77, and 0.71 respectively. The performance of the model in identifying roofs and sport field floorings was lower, with respective F1 scores of 0.57 and 0.59. Overall, the results show that the proposed workflow using deep learning algorithms trained on Sentinel-2 images has a great potential to identify the function of plastic materials on land surfaces.http://www.sciencedirect.com/science/article/pii/S1569843223003151Deep learningEnvironmental managementPlastic detectionPlastic functionalitySentinel-2 |
| spellingShingle | Shanyu Zhou Lichao Mou, Dr. Yuansheng Hua Lixian Zhang Hermann Kaufmann Xiao Xiang Zhu Can we use deep learning models to identify the functionality of plastics from space? International Journal of Applied Earth Observations and Geoinformation Deep learning Environmental management Plastic detection Plastic functionality Sentinel-2 |
| title | Can we use deep learning models to identify the functionality of plastics from space? |
| title_full | Can we use deep learning models to identify the functionality of plastics from space? |
| title_fullStr | Can we use deep learning models to identify the functionality of plastics from space? |
| title_full_unstemmed | Can we use deep learning models to identify the functionality of plastics from space? |
| title_short | Can we use deep learning models to identify the functionality of plastics from space? |
| title_sort | can we use deep learning models to identify the functionality of plastics from space |
| topic | Deep learning Environmental management Plastic detection Plastic functionality Sentinel-2 |
| url | http://www.sciencedirect.com/science/article/pii/S1569843223003151 |
| work_keys_str_mv | AT shanyuzhou canweusedeeplearningmodelstoidentifythefunctionalityofplasticsfromspace AT lichaomoudr canweusedeeplearningmodelstoidentifythefunctionalityofplasticsfromspace AT yuanshenghua canweusedeeplearningmodelstoidentifythefunctionalityofplasticsfromspace AT lixianzhang canweusedeeplearningmodelstoidentifythefunctionalityofplasticsfromspace AT hermannkaufmann canweusedeeplearningmodelstoidentifythefunctionalityofplasticsfromspace AT xiaoxiangzhu canweusedeeplearningmodelstoidentifythefunctionalityofplasticsfromspace |