A high-resolution remote sensing land use/land cover classification method based on multi-level features adaptation of segment anything model
Land use/land cover (LULC) classification based on deep learning techniques is a significant research area for analyzing high-resolution remote sensing(HRRS) images. However, due to the limitation of available samples and model feature extraction capability, the current deep learning methods suffer...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-07-01
|
| Series: | International Journal of Applied Earth Observations and Geoinformation |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1569843225003061 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850116143914680320 |
|---|---|
| author | Hui Yang Zhipeng Jiang Yaobo Zhang Yanlan Wu Heng Luo Peng Zhang Biao Wang |
| author_facet | Hui Yang Zhipeng Jiang Yaobo Zhang Yanlan Wu Heng Luo Peng Zhang Biao Wang |
| author_sort | Hui Yang |
| collection | DOAJ |
| description | Land use/land cover (LULC) classification based on deep learning techniques is a significant research area for analyzing high-resolution remote sensing(HRRS) images. However, due to the limitation of available samples and model feature extraction capability, the current deep learning methods suffer from weak generalization ability for widespread and effective application across diverse HRRS scenarios. To address this problem, we propose an innovative network model named multi-level feature adaptation-segment anything Model (MLFA-SAM). The model employs a three-level fine-tuning strategy to adapt the SAM foundation model for remote sensing LULC classification. The proposed MLFA-SAM significantly enhances high-precision classification performance across diverse HRRS scenarios. Specifically, the domain distribution shift adaptation (DDSA) level is designed to adjust the input image modality for SAM and initially extract features and overcome the domain distribution shift between remote sensing images and the natural images used by the SAM. Then, we designed depthwise low-rank adaptation (DLRA) strategy to optimally fine-tune the frozen SAM parameters. Finally, we improved SAM’s mask decoder to generate high-quality multi-class masks required for LULC classification. Experimental results demonstrate that the MLFA-SAM model surpasses several existing state-of-the-art(SOTA) methods on the HRLC dataset and the ISPRS Potsdam dataset. Quantitative evaluations demonstrate that MLFA-SAM, with its concise yet efficient architecture, achieves 66.77% mIoU and 86.02% OA on the HRLC dataset. Notably, the integration of near-infrared (Nir) bands further enhances its performance to 68.43% mIoU and 87.91% OA. The generalization test on the LoveDA dataset, along with four test HRRS images exhibiting spatiotemporal and semantic scene differences, further demonstrate that MLFA-SAM possesses a stronger generalization ability compared to existing methods and shows greater potential for practical applications. |
| format | Article |
| id | doaj-art-428e575dcf044586a8ee6a2af5e5e573 |
| institution | OA Journals |
| issn | 1569-8432 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Elsevier |
| record_format | Article |
| series | International Journal of Applied Earth Observations and Geoinformation |
| spelling | doaj-art-428e575dcf044586a8ee6a2af5e5e5732025-08-20T02:36:23ZengElsevierInternational Journal of Applied Earth Observations and Geoinformation1569-84322025-07-0114110465910.1016/j.jag.2025.104659A high-resolution remote sensing land use/land cover classification method based on multi-level features adaptation of segment anything modelHui Yang0Zhipeng Jiang1Yaobo Zhang2Yanlan Wu3Heng Luo4Peng Zhang5Biao Wang6Institutes of Physical Science and Information Technology, Anhui University, Hefei 230601, ChinaSchool of Resources and Environmental Engineering, Anhui University, Hefei 230601, ChinaAnhui Land Satellite Remote Sensing Application Technology Center, Hefei 230031, ChinaSchool of Artificial Intelligence, Anhui University, Hefei 230601, China; Corresponding author.Guangxi Zhuang Automomous Region Institute of Natural Resources Remote Sensing, Nanning 530023, ChinaSchool of Artificial Intelligence, Anhui University, Hefei 230601, ChinaSchool of Resources and Environmental Engineering, Anhui University, Hefei 230601, ChinaLand use/land cover (LULC) classification based on deep learning techniques is a significant research area for analyzing high-resolution remote sensing(HRRS) images. However, due to the limitation of available samples and model feature extraction capability, the current deep learning methods suffer from weak generalization ability for widespread and effective application across diverse HRRS scenarios. To address this problem, we propose an innovative network model named multi-level feature adaptation-segment anything Model (MLFA-SAM). The model employs a three-level fine-tuning strategy to adapt the SAM foundation model for remote sensing LULC classification. The proposed MLFA-SAM significantly enhances high-precision classification performance across diverse HRRS scenarios. Specifically, the domain distribution shift adaptation (DDSA) level is designed to adjust the input image modality for SAM and initially extract features and overcome the domain distribution shift between remote sensing images and the natural images used by the SAM. Then, we designed depthwise low-rank adaptation (DLRA) strategy to optimally fine-tune the frozen SAM parameters. Finally, we improved SAM’s mask decoder to generate high-quality multi-class masks required for LULC classification. Experimental results demonstrate that the MLFA-SAM model surpasses several existing state-of-the-art(SOTA) methods on the HRLC dataset and the ISPRS Potsdam dataset. Quantitative evaluations demonstrate that MLFA-SAM, with its concise yet efficient architecture, achieves 66.77% mIoU and 86.02% OA on the HRLC dataset. Notably, the integration of near-infrared (Nir) bands further enhances its performance to 68.43% mIoU and 87.91% OA. The generalization test on the LoveDA dataset, along with four test HRRS images exhibiting spatiotemporal and semantic scene differences, further demonstrate that MLFA-SAM possesses a stronger generalization ability compared to existing methods and shows greater potential for practical applications.http://www.sciencedirect.com/science/article/pii/S1569843225003061Land use/land cover classificationSegment anything model(SAM)Parameter efficient fine-tuningHigh-resolution remote sensing imagesSemantic segmentation |
| spellingShingle | Hui Yang Zhipeng Jiang Yaobo Zhang Yanlan Wu Heng Luo Peng Zhang Biao Wang A high-resolution remote sensing land use/land cover classification method based on multi-level features adaptation of segment anything model International Journal of Applied Earth Observations and Geoinformation Land use/land cover classification Segment anything model(SAM) Parameter efficient fine-tuning High-resolution remote sensing images Semantic segmentation |
| title | A high-resolution remote sensing land use/land cover classification method based on multi-level features adaptation of segment anything model |
| title_full | A high-resolution remote sensing land use/land cover classification method based on multi-level features adaptation of segment anything model |
| title_fullStr | A high-resolution remote sensing land use/land cover classification method based on multi-level features adaptation of segment anything model |
| title_full_unstemmed | A high-resolution remote sensing land use/land cover classification method based on multi-level features adaptation of segment anything model |
| title_short | A high-resolution remote sensing land use/land cover classification method based on multi-level features adaptation of segment anything model |
| title_sort | high resolution remote sensing land use land cover classification method based on multi level features adaptation of segment anything model |
| topic | Land use/land cover classification Segment anything model(SAM) Parameter efficient fine-tuning High-resolution remote sensing images Semantic segmentation |
| url | http://www.sciencedirect.com/science/article/pii/S1569843225003061 |
| work_keys_str_mv | AT huiyang ahighresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT zhipengjiang ahighresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT yaobozhang ahighresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT yanlanwu ahighresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT hengluo ahighresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT pengzhang ahighresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT biaowang ahighresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT huiyang highresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT zhipengjiang highresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT yaobozhang highresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT yanlanwu highresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT hengluo highresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT pengzhang highresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel AT biaowang highresolutionremotesensinglanduselandcoverclassificationmethodbasedonmultilevelfeaturesadaptationofsegmentanythingmodel |