SyntCities: A Large Synthetic Remote Sensing Dataset for Disparity Estimation
Studies in the last years have proved the outstanding performance of deep learning for computer vision tasks in the remote sensing field, such as disparity estimation. However, available datasets mostly focus on close-range applications like autonomous driving or robot manipulation. To reduce the do...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9960780/ |
| _version_ | 1811212381213163520 |
|---|---|
| author | Mario Fuentes Reyes Pablo D'Angelo Friedrich Fraundorfer |
| author_facet | Mario Fuentes Reyes Pablo D'Angelo Friedrich Fraundorfer |
| author_sort | Mario Fuentes Reyes |
| collection | DOAJ |
| description | Studies in the last years have proved the outstanding performance of deep learning for computer vision tasks in the remote sensing field, such as disparity estimation. However, available datasets mostly focus on close-range applications like autonomous driving or robot manipulation. To reduce the domain gap while training we present SyntCities, a synthetic dataset resembling the aerial imagery on urban areas. The pipeline used to render the images is based on 3-D modeling, which helps to avoid acquisition costs, provides subpixel accurate dense ground truth and simulates different illumination conditions. The dataset additionally provides multiclass semantic maps and can be converted to point cloud format to benefit a wider research community. We focus on the task of disparity estimation and evaluate the performance of the traditional semiglobal matching and state-of-the-art architectures, trained with SyntCities and other datasets, on real aerial and satellite images. A comparison with the widely used SceneFlow dataset is also presented. Strategies using a mixture of both real and synthetic samples are studied as well. Results show significant improvements in terms of accuracy for the disparity maps. |
| first_indexed | 2024-04-12T05:28:30Z |
| format | Article |
| id | doaj.art-d6fc3f5c6ffe487dbd72af197e6f50c5 |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-04-12T05:28:30Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-d6fc3f5c6ffe487dbd72af197e6f50c52022-12-22T03:46:11ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352022-01-0115100871009810.1109/JSTARS.2022.32239379960780SyntCities: A Large Synthetic Remote Sensing Dataset for Disparity EstimationMario Fuentes Reyes0https://orcid.org/0000-0002-6593-5152Pablo D'Angelo1https://orcid.org/0000-0001-8541-3856Friedrich Fraundorfer2https://orcid.org/0000-0002-5805-8892Department of Photogrammetry and Image Analysis, Remote Sensing Technology Institute (IMF), German Aerospace Center (DLR), Wessling, GermanyDepartment of Photogrammetry and Image Analysis, Remote Sensing Technology Institute (IMF), German Aerospace Center (DLR), Wessling, GermanyDepartment of Photogrammetry and Image Analysis, Remote Sensing Technology Institute (IMF), German Aerospace Center (DLR), Wessling, GermanyStudies in the last years have proved the outstanding performance of deep learning for computer vision tasks in the remote sensing field, such as disparity estimation. However, available datasets mostly focus on close-range applications like autonomous driving or robot manipulation. To reduce the domain gap while training we present SyntCities, a synthetic dataset resembling the aerial imagery on urban areas. The pipeline used to render the images is based on 3-D modeling, which helps to avoid acquisition costs, provides subpixel accurate dense ground truth and simulates different illumination conditions. The dataset additionally provides multiclass semantic maps and can be converted to point cloud format to benefit a wider research community. We focus on the task of disparity estimation and evaluate the performance of the traditional semiglobal matching and state-of-the-art architectures, trained with SyntCities and other datasets, on real aerial and satellite images. A comparison with the widely used SceneFlow dataset is also presented. Strategies using a mixture of both real and synthetic samples are studied as well. Results show significant improvements in terms of accuracy for the disparity maps.https://ieeexplore.ieee.org/document/9960780/Disparity estimationsynthetic imageryurban reconstruction |
| spellingShingle | Mario Fuentes Reyes Pablo D'Angelo Friedrich Fraundorfer SyntCities: A Large Synthetic Remote Sensing Dataset for Disparity Estimation IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Disparity estimation synthetic imagery urban reconstruction |
| title | SyntCities: A Large Synthetic Remote Sensing Dataset for Disparity Estimation |
| title_full | SyntCities: A Large Synthetic Remote Sensing Dataset for Disparity Estimation |
| title_fullStr | SyntCities: A Large Synthetic Remote Sensing Dataset for Disparity Estimation |
| title_full_unstemmed | SyntCities: A Large Synthetic Remote Sensing Dataset for Disparity Estimation |
| title_short | SyntCities: A Large Synthetic Remote Sensing Dataset for Disparity Estimation |
| title_sort | syntcities a large synthetic remote sensing dataset for disparity estimation |
| topic | Disparity estimation synthetic imagery urban reconstruction |
| url | https://ieeexplore.ieee.org/document/9960780/ |
| work_keys_str_mv | AT mariofuentesreyes syntcitiesalargesyntheticremotesensingdatasetfordisparityestimation AT pablodangelo syntcitiesalargesyntheticremotesensingdatasetfordisparityestimation AT friedrichfraundorfer syntcitiesalargesyntheticremotesensingdatasetfordisparityestimation |