MCDNet: Multilevel cloud detection network for remote sensing images based on dual-perspective change-guided and multi-scale feature fusion
Cloud detection plays a crucial role in the preprocessing of optical remote sensing images. While extensive deep learning-based methods have shown strong performance in detecting thick clouds, their ability to identify thin and broken clouds is often inadequate due to their sparse distribution, semi...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2024-05-01
|
| Series: | International Journal of Applied Earth Observations and Geoinformation |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1569843224001742 |
| _version_ | 1797213413987844096 |
|---|---|
| author | Junwu Dong Yanhui Wang Yang Yang Mengqin Yang Jun Chen |
| author_facet | Junwu Dong Yanhui Wang Yang Yang Mengqin Yang Jun Chen |
| author_sort | Junwu Dong |
| collection | DOAJ |
| description | Cloud detection plays a crucial role in the preprocessing of optical remote sensing images. While extensive deep learning-based methods have shown strong performance in detecting thick clouds, their ability to identify thin and broken clouds is often inadequate due to their sparse distribution, semi-transparency, and similarity to background regions. To address this limitation, we introduce a multilevel cloud detection network (MCDNet) capable of simultaneously detecting thick and thin clouds. This network effectively enhances the accuracy of identifying thin and broken clouds by integrating a dual-perspective change-guided mechanism (DPCG) and a multi-scale feature fusion module (MSFF). The DPCG creates a dual-input stream by combining the original image with the thin cloud removal image, and then utilizes a dual-perspective feature fusion module (DPFF) to perform feature fusion and extract change features, thereby improving the model's ability to perceive thin cloud regions and mitigate inter-class similarity in multilevel cloud detection. The MSFF enhances the model's sensitivity to broken clouds by utilizing multiple non-adjacent low-level features to remedy the missing spatial information in the high-level features during multiple downsampling. Experimental results on the L8-Biome and WHUS2-CD datasets demonstrate that MCDNet significantly enhances the detection performance of both thin and broken clouds, and outperforms state-of-the-art methods in accuracy and efficiency. The code of MCDNet is available in https://github.com/djw-easy/MCDNet. |
| first_indexed | 2024-04-24T10:57:54Z |
| format | Article |
| id | doaj.art-d44cb5e75bda4193a9f10d7b8f5422ab |
| institution | Directory Open Access Journal |
| issn | 1569-8432 |
| language | English |
| last_indexed | 2024-04-24T10:57:54Z |
| publishDate | 2024-05-01 |
| publisher | Elsevier |
| record_format | Article |
| series | International Journal of Applied Earth Observations and Geoinformation |
| spelling | doaj.art-d44cb5e75bda4193a9f10d7b8f5422ab2024-04-12T04:45:01ZengElsevierInternational Journal of Applied Earth Observations and Geoinformation1569-84322024-05-01129103820MCDNet: Multilevel cloud detection network for remote sensing images based on dual-perspective change-guided and multi-scale feature fusionJunwu Dong0Yanhui Wang1Yang Yang2Mengqin Yang3Jun Chen4College of Resources Environment and Tourism, Capital Normal University, Beijing 100048, China; Key Laboratory of 3Dimensional Information Acquisition and Application, Ministry of Education, Capital Normal University, Beijing 100048, ChinaCollege of Resources Environment and Tourism, Capital Normal University, Beijing 100048, China; Key Laboratory of 3Dimensional Information Acquisition and Application, Ministry of Education, Capital Normal University, Beijing 100048, China; Corresponding author at: College of Resources Environment and Tourism, Capital Normal University, No.105, West Third Ring Road, Haidian District, Beijing 100048, China.College of Resources Environment and Tourism, Capital Normal University, Beijing 100048, China; Key Laboratory of 3Dimensional Information Acquisition and Application, Ministry of Education, Capital Normal University, Beijing 100048, ChinaCollege of Resources Environment and Tourism, Capital Normal University, Beijing 100048, China; Key Laboratory of 3Dimensional Information Acquisition and Application, Ministry of Education, Capital Normal University, Beijing 100048, ChinaSchool of Environment Science and Spatial Informatics, China University of Mining and Technology, Xuzhou 221116, China; National Geomatics Center of China, Beijing 100830, China; Moganshan Geospatial Information Lab, Deqing 313299, ChinaCloud detection plays a crucial role in the preprocessing of optical remote sensing images. While extensive deep learning-based methods have shown strong performance in detecting thick clouds, their ability to identify thin and broken clouds is often inadequate due to their sparse distribution, semi-transparency, and similarity to background regions. To address this limitation, we introduce a multilevel cloud detection network (MCDNet) capable of simultaneously detecting thick and thin clouds. This network effectively enhances the accuracy of identifying thin and broken clouds by integrating a dual-perspective change-guided mechanism (DPCG) and a multi-scale feature fusion module (MSFF). The DPCG creates a dual-input stream by combining the original image with the thin cloud removal image, and then utilizes a dual-perspective feature fusion module (DPFF) to perform feature fusion and extract change features, thereby improving the model's ability to perceive thin cloud regions and mitigate inter-class similarity in multilevel cloud detection. The MSFF enhances the model's sensitivity to broken clouds by utilizing multiple non-adjacent low-level features to remedy the missing spatial information in the high-level features during multiple downsampling. Experimental results on the L8-Biome and WHUS2-CD datasets demonstrate that MCDNet significantly enhances the detection performance of both thin and broken clouds, and outperforms state-of-the-art methods in accuracy and efficiency. The code of MCDNet is available in https://github.com/djw-easy/MCDNet.http://www.sciencedirect.com/science/article/pii/S1569843224001742Cloud detectionRemote sensingDual-perspective change-guidedMulti-scale feature fusion |
| spellingShingle | Junwu Dong Yanhui Wang Yang Yang Mengqin Yang Jun Chen MCDNet: Multilevel cloud detection network for remote sensing images based on dual-perspective change-guided and multi-scale feature fusion International Journal of Applied Earth Observations and Geoinformation Cloud detection Remote sensing Dual-perspective change-guided Multi-scale feature fusion |
| title | MCDNet: Multilevel cloud detection network for remote sensing images based on dual-perspective change-guided and multi-scale feature fusion |
| title_full | MCDNet: Multilevel cloud detection network for remote sensing images based on dual-perspective change-guided and multi-scale feature fusion |
| title_fullStr | MCDNet: Multilevel cloud detection network for remote sensing images based on dual-perspective change-guided and multi-scale feature fusion |
| title_full_unstemmed | MCDNet: Multilevel cloud detection network for remote sensing images based on dual-perspective change-guided and multi-scale feature fusion |
| title_short | MCDNet: Multilevel cloud detection network for remote sensing images based on dual-perspective change-guided and multi-scale feature fusion |
| title_sort | mcdnet multilevel cloud detection network for remote sensing images based on dual perspective change guided and multi scale feature fusion |
| topic | Cloud detection Remote sensing Dual-perspective change-guided Multi-scale feature fusion |
| url | http://www.sciencedirect.com/science/article/pii/S1569843224001742 |
| work_keys_str_mv | AT junwudong mcdnetmultilevelclouddetectionnetworkforremotesensingimagesbasedondualperspectivechangeguidedandmultiscalefeaturefusion AT yanhuiwang mcdnetmultilevelclouddetectionnetworkforremotesensingimagesbasedondualperspectivechangeguidedandmultiscalefeaturefusion AT yangyang mcdnetmultilevelclouddetectionnetworkforremotesensingimagesbasedondualperspectivechangeguidedandmultiscalefeaturefusion AT mengqinyang mcdnetmultilevelclouddetectionnetworkforremotesensingimagesbasedondualperspectivechangeguidedandmultiscalefeaturefusion AT junchen mcdnetmultilevelclouddetectionnetworkforremotesensingimagesbasedondualperspectivechangeguidedandmultiscalefeaturefusion |