Shadow Detection and Reconstruction of High-Resolution Remote Sensing Images in Mountainous and Hilly Environments
The undulating terrain in mountainous and hilly regions results in a greater variety and complexity of shadows. Efficient methods for shadow detection and reconstruction in high-resolution remote sensing images are particularly important in such hilly areas. The accurate detection of shadow masks is...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10339831/ |
| _version_ | 1827399032034557952 |
|---|---|
| author | Zhenqing Wang Yi Zhou Futao Wang Shixin Wang Gang Qin Jinfeng Zhu |
| author_facet | Zhenqing Wang Yi Zhou Futao Wang Shixin Wang Gang Qin Jinfeng Zhu |
| author_sort | Zhenqing Wang |
| collection | DOAJ |
| description | The undulating terrain in mountainous and hilly regions results in a greater variety and complexity of shadows. Efficient methods for shadow detection and reconstruction in high-resolution remote sensing images are particularly important in such hilly areas. The accurate detection of shadow masks is a prerequisite for shadow reconstruction. By utilizing the features of high hue and low intensity in shadow areas, an initial spectral ratio is constructed based on the CIELCh color space model. Simple linear iterative clustering is employed to perform superpixel segmentation on the image, and the segmented results are spatially constrained to reconstruct the initial spectral ratio. Afterward, an automatic multilevel global thresholding approach is applied to obtain the shadow mask and eliminate the influence of interfering objects. For shadow reconstruction, the segmented superpixels are treated as the smallest processing units. Similar neighboring objects have similar ambient light intensities. Based on this, we propose a shadow reconstruction method, which compensates shadow superpixels using adjacent nonshadow superpixels and determines compensation weights based on their similarity. Furthermore, the shadow boundaries are dilated to obtain penumbra, and mean filtering is performed to compensate for the illumination in the penumbra. Finally, the proposed method is qualitatively and quantitatively compared with existing shadow detection and reconstruction methods. Experimental results demonstrate that this method can accurately detect shadows in high-resolution remote sensing images in mountainous and hilly environments, and effectively reconstruct the spectral information of shadow areas. This has significant implications for subsequent feature extraction and further analysis in mountainous and hilly regions. |
| first_indexed | 2024-03-08T19:38:39Z |
| format | Article |
| id | doaj.art-9eba49b9d02349d18ca6b48d8b7cdceb |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-03-08T19:38:39Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-9eba49b9d02349d18ca6b48d8b7cdceb2023-12-26T00:01:15ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352024-01-01171233124310.1109/JSTARS.2023.333897610339831Shadow Detection and Reconstruction of High-Resolution Remote Sensing Images in Mountainous and Hilly EnvironmentsZhenqing Wang0https://orcid.org/0000-0001-6394-2458Yi Zhou1Futao Wang2https://orcid.org/0000-0002-0179-0214Shixin Wang3Gang Qin4https://orcid.org/0000-0002-0024-3251Jinfeng Zhu5Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaThe undulating terrain in mountainous and hilly regions results in a greater variety and complexity of shadows. Efficient methods for shadow detection and reconstruction in high-resolution remote sensing images are particularly important in such hilly areas. The accurate detection of shadow masks is a prerequisite for shadow reconstruction. By utilizing the features of high hue and low intensity in shadow areas, an initial spectral ratio is constructed based on the CIELCh color space model. Simple linear iterative clustering is employed to perform superpixel segmentation on the image, and the segmented results are spatially constrained to reconstruct the initial spectral ratio. Afterward, an automatic multilevel global thresholding approach is applied to obtain the shadow mask and eliminate the influence of interfering objects. For shadow reconstruction, the segmented superpixels are treated as the smallest processing units. Similar neighboring objects have similar ambient light intensities. Based on this, we propose a shadow reconstruction method, which compensates shadow superpixels using adjacent nonshadow superpixels and determines compensation weights based on their similarity. Furthermore, the shadow boundaries are dilated to obtain penumbra, and mean filtering is performed to compensate for the illumination in the penumbra. Finally, the proposed method is qualitatively and quantitatively compared with existing shadow detection and reconstruction methods. Experimental results demonstrate that this method can accurately detect shadows in high-resolution remote sensing images in mountainous and hilly environments, and effectively reconstruct the spectral information of shadow areas. This has significant implications for subsequent feature extraction and further analysis in mountainous and hilly regions.https://ieeexplore.ieee.org/document/10339831/High-resolution remote sensinghilly environmentshadow detectionshadow reconstruction |
| spellingShingle | Zhenqing Wang Yi Zhou Futao Wang Shixin Wang Gang Qin Jinfeng Zhu Shadow Detection and Reconstruction of High-Resolution Remote Sensing Images in Mountainous and Hilly Environments IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing High-resolution remote sensing hilly environment shadow detection shadow reconstruction |
| title | Shadow Detection and Reconstruction of High-Resolution Remote Sensing Images in Mountainous and Hilly Environments |
| title_full | Shadow Detection and Reconstruction of High-Resolution Remote Sensing Images in Mountainous and Hilly Environments |
| title_fullStr | Shadow Detection and Reconstruction of High-Resolution Remote Sensing Images in Mountainous and Hilly Environments |
| title_full_unstemmed | Shadow Detection and Reconstruction of High-Resolution Remote Sensing Images in Mountainous and Hilly Environments |
| title_short | Shadow Detection and Reconstruction of High-Resolution Remote Sensing Images in Mountainous and Hilly Environments |
| title_sort | shadow detection and reconstruction of high resolution remote sensing images in mountainous and hilly environments |
| topic | High-resolution remote sensing hilly environment shadow detection shadow reconstruction |
| url | https://ieeexplore.ieee.org/document/10339831/ |
| work_keys_str_mv | AT zhenqingwang shadowdetectionandreconstructionofhighresolutionremotesensingimagesinmountainousandhillyenvironments AT yizhou shadowdetectionandreconstructionofhighresolutionremotesensingimagesinmountainousandhillyenvironments AT futaowang shadowdetectionandreconstructionofhighresolutionremotesensingimagesinmountainousandhillyenvironments AT shixinwang shadowdetectionandreconstructionofhighresolutionremotesensingimagesinmountainousandhillyenvironments AT gangqin shadowdetectionandreconstructionofhighresolutionremotesensingimagesinmountainousandhillyenvironments AT jinfengzhu shadowdetectionandreconstructionofhighresolutionremotesensingimagesinmountainousandhillyenvironments |