A Refined Terrace Extraction Method Based on a Local Optimization Model Using GF-2 Images
Terraces are an important form of surface modification, and their spatial distribution data are of utmost importance for ensuring food and water security. However, the extraction of terrace patches faces challenges due to the complexity of the terrain and limitations in remote sensing (RS) data. The...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/17/1/12 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841548997361139712 |
|---|---|
| author | Guobin Kan Jie Gong Bao Wang Xia Li Jing Shi Yutao Ma Wei Wei Jun Zhang |
| author_facet | Guobin Kan Jie Gong Bao Wang Xia Li Jing Shi Yutao Ma Wei Wei Jun Zhang |
| author_sort | Guobin Kan |
| collection | DOAJ |
| description | Terraces are an important form of surface modification, and their spatial distribution data are of utmost importance for ensuring food and water security. However, the extraction of terrace patches faces challenges due to the complexity of the terrain and limitations in remote sensing (RS) data. Therefore, there is an urgent need for advanced technology models that can accurately extract terraces. High-resolution RS data allows for detailed characterization of terraces by capturing more precise surface features. Moreover, leveraging deep learning (DL) models with local adaptive improvements can further enhance the accuracy of interpretation by exploring latent information. In this study, we employed five models: ResU-Net, U-Net++, RVTransUNet, XDeepLabV3+, and ResPSPNet as DL models to extract fine patch terraces from GF-2 images. We then integrated morphological, textural, and spectral features to optimize the extraction process by addressing issues related to low adhesion and edge segmentation performance. The model structure and loss function were adjusted accordingly to achieve high-quality terrace mapping results. Finally, we utilized multi-source RS data along with terrain elements for correction and optimization to generate a 1 m resolution terrace distribution map in the Zuli River Basin (TDZRB). Evaluation results after correction demonstrate that our approach achieved an OA, F1-Score, and MIoU of 96.67%, 93.94%, and 89.37%, respectively. The total area of terraces in the Zuli River Basin was calculated at 2557 ± 117.96 km<sup>2</sup> using EM with our model methodology; this accounts for approximately 41.74% ± 1.93% of the cultivated land area within the Zuli River Basin. Therefore, obtaining accurate information on patch terrace distribution serves as essential foundational data for terrace ecosystem research and government decision-making. |
| format | Article |
| id | doaj-art-6ff996f8c1e3400d818bb01be5c94f94 |
| institution | Kabale University |
| issn | 2072-4292 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj-art-6ff996f8c1e3400d818bb01be5c94f942025-01-10T13:19:56ZengMDPI AGRemote Sensing2072-42922024-12-011711210.3390/rs17010012A Refined Terrace Extraction Method Based on a Local Optimization Model Using GF-2 ImagesGuobin Kan0Jie Gong1Bao Wang2Xia Li3Jing Shi4Yutao Ma5Wei Wei6Jun Zhang7College of Earth and Environmental Sciences, Lanzhou University, Lanzhou 730000, ChinaCollege of Earth and Environmental Sciences, Lanzhou University, Lanzhou 730000, ChinaCollege of Earth and Environmental Sciences, Lanzhou University, Lanzhou 730000, ChinaCollege of Earth and Environmental Sciences, Lanzhou University, Lanzhou 730000, ChinaCollege of Earth and Environmental Sciences, Lanzhou University, Lanzhou 730000, ChinaCollege of Earth and Environmental Sciences, Lanzhou University, Lanzhou 730000, ChinaState Key Laboratory of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, ChinaGansu Academy of Eco-Environmental Sciences, Lanzhou 730000, ChinaTerraces are an important form of surface modification, and their spatial distribution data are of utmost importance for ensuring food and water security. However, the extraction of terrace patches faces challenges due to the complexity of the terrain and limitations in remote sensing (RS) data. Therefore, there is an urgent need for advanced technology models that can accurately extract terraces. High-resolution RS data allows for detailed characterization of terraces by capturing more precise surface features. Moreover, leveraging deep learning (DL) models with local adaptive improvements can further enhance the accuracy of interpretation by exploring latent information. In this study, we employed five models: ResU-Net, U-Net++, RVTransUNet, XDeepLabV3+, and ResPSPNet as DL models to extract fine patch terraces from GF-2 images. We then integrated morphological, textural, and spectral features to optimize the extraction process by addressing issues related to low adhesion and edge segmentation performance. The model structure and loss function were adjusted accordingly to achieve high-quality terrace mapping results. Finally, we utilized multi-source RS data along with terrain elements for correction and optimization to generate a 1 m resolution terrace distribution map in the Zuli River Basin (TDZRB). Evaluation results after correction demonstrate that our approach achieved an OA, F1-Score, and MIoU of 96.67%, 93.94%, and 89.37%, respectively. The total area of terraces in the Zuli River Basin was calculated at 2557 ± 117.96 km<sup>2</sup> using EM with our model methodology; this accounts for approximately 41.74% ± 1.93% of the cultivated land area within the Zuli River Basin. Therefore, obtaining accurate information on patch terrace distribution serves as essential foundational data for terrace ecosystem research and government decision-making.https://www.mdpi.com/2072-4292/17/1/12deep learningGF-2terrace extractionremote sensing mapping |
| spellingShingle | Guobin Kan Jie Gong Bao Wang Xia Li Jing Shi Yutao Ma Wei Wei Jun Zhang A Refined Terrace Extraction Method Based on a Local Optimization Model Using GF-2 Images Remote Sensing deep learning GF-2 terrace extraction remote sensing mapping |
| title | A Refined Terrace Extraction Method Based on a Local Optimization Model Using GF-2 Images |
| title_full | A Refined Terrace Extraction Method Based on a Local Optimization Model Using GF-2 Images |
| title_fullStr | A Refined Terrace Extraction Method Based on a Local Optimization Model Using GF-2 Images |
| title_full_unstemmed | A Refined Terrace Extraction Method Based on a Local Optimization Model Using GF-2 Images |
| title_short | A Refined Terrace Extraction Method Based on a Local Optimization Model Using GF-2 Images |
| title_sort | refined terrace extraction method based on a local optimization model using gf 2 images |
| topic | deep learning GF-2 terrace extraction remote sensing mapping |
| url | https://www.mdpi.com/2072-4292/17/1/12 |
| work_keys_str_mv | AT guobinkan arefinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT jiegong arefinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT baowang arefinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT xiali arefinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT jingshi arefinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT yutaoma arefinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT weiwei arefinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT junzhang arefinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT guobinkan refinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT jiegong refinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT baowang refinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT xiali refinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT jingshi refinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT yutaoma refinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT weiwei refinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images AT junzhang refinedterraceextractionmethodbasedonalocaloptimizationmodelusinggf2images |