Controllably Deep Supervision and Multi-Scale Feature Fusion Network for Cloud and Snow Detection Based on Medium- and High-Resolution Imagery Dataset
Clouds and snow in remote sensing imageries cover underlying surface information, reducing image availability. Moreover, they interact with each other, decreasing the cloud and snow detection accuracy. In this study, we propose a convolutional neural network for cloud and snow detection, named the c...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-11-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/23/4805 |
| _version_ | 1797507261334028288 |
|---|---|
| author | Guangbin Zhang Xianjun Gao Yuanwei Yang Mingwei Wang Shuhao Ran |
| author_facet | Guangbin Zhang Xianjun Gao Yuanwei Yang Mingwei Wang Shuhao Ran |
| author_sort | Guangbin Zhang |
| collection | DOAJ |
| description | Clouds and snow in remote sensing imageries cover underlying surface information, reducing image availability. Moreover, they interact with each other, decreasing the cloud and snow detection accuracy. In this study, we propose a convolutional neural network for cloud and snow detection, named the cloud and snow detection network (CSD-Net). It incorporates the multi-scale feature fusion module (MFF) and the controllably deep supervision and feature fusion structure (CDSFF). MFF can capture and aggregate features at various scales, ensuring that the extracted high-level semantic features of clouds and snow are more distinctive. CDSFF can provide a deeply supervised mechanism with hinge loss and combine information from adjacent layers to gain more representative features. It ensures the gradient flow is more oriented and error-less, while retaining more effective information. Additionally, a high-resolution cloud and snow dataset based on WorldView2 (CSWV) was created and released. This dataset meets the training requirements of deep learning methods for clouds and snow in high-resolution remote sensing images. Based on the datasets with varied resolutions, CSD-Net is compared to eight state-of-the-art deep learning methods. The experiment results indicate that CSD-Net has an excellent detection accuracy and efficiency. Specifically, the mean intersection over the union (MIoU) of CSD-Net is the highest in the corresponding experiment. Furthermore, the number of parameters in our proposed network is just 7.61 million, which is the lowest of the tested methods. It only has 88.06 GFLOPs of floating point operations, which is less than the U-Net, DeepLabV3+, PSPNet, SegNet-Modified, MSCFF, and GeoInfoNet. Meanwhile, CSWV has a higher annotation quality since the same method can obtain a greater accuracy on it. |
| first_indexed | 2024-03-10T04:46:01Z |
| format | Article |
| id | doaj.art-049c8ed413be4eccaa88a6319500c3cc |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T04:46:01Z |
| publishDate | 2021-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-049c8ed413be4eccaa88a6319500c3cc2023-11-23T02:56:37ZengMDPI AGRemote Sensing2072-42922021-11-011323480510.3390/rs13234805Controllably Deep Supervision and Multi-Scale Feature Fusion Network for Cloud and Snow Detection Based on Medium- and High-Resolution Imagery DatasetGuangbin Zhang0Xianjun Gao1Yuanwei Yang2Mingwei Wang3Shuhao Ran4School of Geosciences, Yangtze University, Wuhan 430100, ChinaSchool of Geosciences, Yangtze University, Wuhan 430100, ChinaSchool of Geosciences, Yangtze University, Wuhan 430100, ChinaSchool of Computer Science, Hubei University of Technology, Wuhan 430068, ChinaSchool of Geosciences, Yangtze University, Wuhan 430100, ChinaClouds and snow in remote sensing imageries cover underlying surface information, reducing image availability. Moreover, they interact with each other, decreasing the cloud and snow detection accuracy. In this study, we propose a convolutional neural network for cloud and snow detection, named the cloud and snow detection network (CSD-Net). It incorporates the multi-scale feature fusion module (MFF) and the controllably deep supervision and feature fusion structure (CDSFF). MFF can capture and aggregate features at various scales, ensuring that the extracted high-level semantic features of clouds and snow are more distinctive. CDSFF can provide a deeply supervised mechanism with hinge loss and combine information from adjacent layers to gain more representative features. It ensures the gradient flow is more oriented and error-less, while retaining more effective information. Additionally, a high-resolution cloud and snow dataset based on WorldView2 (CSWV) was created and released. This dataset meets the training requirements of deep learning methods for clouds and snow in high-resolution remote sensing images. Based on the datasets with varied resolutions, CSD-Net is compared to eight state-of-the-art deep learning methods. The experiment results indicate that CSD-Net has an excellent detection accuracy and efficiency. Specifically, the mean intersection over the union (MIoU) of CSD-Net is the highest in the corresponding experiment. Furthermore, the number of parameters in our proposed network is just 7.61 million, which is the lowest of the tested methods. It only has 88.06 GFLOPs of floating point operations, which is less than the U-Net, DeepLabV3+, PSPNet, SegNet-Modified, MSCFF, and GeoInfoNet. Meanwhile, CSWV has a higher annotation quality since the same method can obtain a greater accuracy on it.https://www.mdpi.com/2072-4292/13/23/4805cloud and snow detectionconvolutional neural networkcontrollably deep supervisionmulti-scale feature fusioncloud and snow dataset |
| spellingShingle | Guangbin Zhang Xianjun Gao Yuanwei Yang Mingwei Wang Shuhao Ran Controllably Deep Supervision and Multi-Scale Feature Fusion Network for Cloud and Snow Detection Based on Medium- and High-Resolution Imagery Dataset Remote Sensing cloud and snow detection convolutional neural network controllably deep supervision multi-scale feature fusion cloud and snow dataset |
| title | Controllably Deep Supervision and Multi-Scale Feature Fusion Network for Cloud and Snow Detection Based on Medium- and High-Resolution Imagery Dataset |
| title_full | Controllably Deep Supervision and Multi-Scale Feature Fusion Network for Cloud and Snow Detection Based on Medium- and High-Resolution Imagery Dataset |
| title_fullStr | Controllably Deep Supervision and Multi-Scale Feature Fusion Network for Cloud and Snow Detection Based on Medium- and High-Resolution Imagery Dataset |
| title_full_unstemmed | Controllably Deep Supervision and Multi-Scale Feature Fusion Network for Cloud and Snow Detection Based on Medium- and High-Resolution Imagery Dataset |
| title_short | Controllably Deep Supervision and Multi-Scale Feature Fusion Network for Cloud and Snow Detection Based on Medium- and High-Resolution Imagery Dataset |
| title_sort | controllably deep supervision and multi scale feature fusion network for cloud and snow detection based on medium and high resolution imagery dataset |
| topic | cloud and snow detection convolutional neural network controllably deep supervision multi-scale feature fusion cloud and snow dataset |
| url | https://www.mdpi.com/2072-4292/13/23/4805 |
| work_keys_str_mv | AT guangbinzhang controllablydeepsupervisionandmultiscalefeaturefusionnetworkforcloudandsnowdetectionbasedonmediumandhighresolutionimagerydataset AT xianjungao controllablydeepsupervisionandmultiscalefeaturefusionnetworkforcloudandsnowdetectionbasedonmediumandhighresolutionimagerydataset AT yuanweiyang controllablydeepsupervisionandmultiscalefeaturefusionnetworkforcloudandsnowdetectionbasedonmediumandhighresolutionimagerydataset AT mingweiwang controllablydeepsupervisionandmultiscalefeaturefusionnetworkforcloudandsnowdetectionbasedonmediumandhighresolutionimagerydataset AT shuhaoran controllablydeepsupervisionandmultiscalefeaturefusionnetworkforcloudandsnowdetectionbasedonmediumandhighresolutionimagerydataset |