Towards semantic segmentation of urban-scale 3D point clouds: A dataset, benchmarks and challenges
An essential prerequisite for unleashing the potential of supervised deep learning algorithms in the area of 3D scene understanding is the availability of large-scale and richly annotated datasets. However, publicly available datasets are either in relative small spatial scales or have limited seman...
| Main Authors: | , , , , , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
IEEE
2021
|
| _version_ | 1826305213166780416 |
|---|---|
| author | Hu, Q Yang, B Khalid, S Xiao, W Trigoni, N Markham, A |
| author_facet | Hu, Q Yang, B Khalid, S Xiao, W Trigoni, N Markham, A |
| author_sort | Hu, Q |
| collection | OXFORD |
| description | An essential prerequisite for unleashing the potential of supervised deep learning algorithms in the area of 3D scene understanding is the availability of large-scale and richly annotated datasets. However, publicly available datasets are either in relative small spatial scales or have limited semantic annotations due to the expensive cost of data acquisition and data annotation, which severely limits the development of fine-grained semantic understanding in the context of 3D point clouds. In this paper, we present an urban-scale photogrammetric point cloud dataset with nearly three billion richly annotated points, which is three times the number of labeled points than the existing largest
photogrammetric point cloud dataset. Our dataset consists of large areas from three UK cities, covering about 7.6 km2 of the city landscape. In the dataset, each 3D point is labeled as one of 13 semantic classes. We extensively evaluate the performance of state-of-the-art algorithms on our dataset and provide a comprehensive analysis of the results. In particular, we identify several key challenges towards urban-scale point cloud understanding. The dataset is available at https://github.com/QingyongHu/SensatUrban.
|
| first_indexed | 2024-03-07T06:29:28Z |
| format | Conference item |
| id | oxford-uuid:f57b7e60-59db-44fb-b3ee-647cfebad98a |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T06:29:28Z |
| publishDate | 2021 |
| publisher | IEEE |
| record_format | dspace |
| spelling | oxford-uuid:f57b7e60-59db-44fb-b3ee-647cfebad98a2022-03-27T12:27:39ZTowards semantic segmentation of urban-scale 3D point clouds: A dataset, benchmarks and challengesConference itemhttp://purl.org/coar/resource_type/c_5794uuid:f57b7e60-59db-44fb-b3ee-647cfebad98aEnglishSymplectic ElementsIEEE2021Hu, QYang, BKhalid, SXiao, WTrigoni, NMarkham, AAn essential prerequisite for unleashing the potential of supervised deep learning algorithms in the area of 3D scene understanding is the availability of large-scale and richly annotated datasets. However, publicly available datasets are either in relative small spatial scales or have limited semantic annotations due to the expensive cost of data acquisition and data annotation, which severely limits the development of fine-grained semantic understanding in the context of 3D point clouds. In this paper, we present an urban-scale photogrammetric point cloud dataset with nearly three billion richly annotated points, which is three times the number of labeled points than the existing largest photogrammetric point cloud dataset. Our dataset consists of large areas from three UK cities, covering about 7.6 km2 of the city landscape. In the dataset, each 3D point is labeled as one of 13 semantic classes. We extensively evaluate the performance of state-of-the-art algorithms on our dataset and provide a comprehensive analysis of the results. In particular, we identify several key challenges towards urban-scale point cloud understanding. The dataset is available at https://github.com/QingyongHu/SensatUrban. |
| spellingShingle | Hu, Q Yang, B Khalid, S Xiao, W Trigoni, N Markham, A Towards semantic segmentation of urban-scale 3D point clouds: A dataset, benchmarks and challenges |
| title | Towards semantic segmentation of urban-scale 3D point clouds: A dataset, benchmarks and challenges |
| title_full | Towards semantic segmentation of urban-scale 3D point clouds: A dataset, benchmarks and challenges |
| title_fullStr | Towards semantic segmentation of urban-scale 3D point clouds: A dataset, benchmarks and challenges |
| title_full_unstemmed | Towards semantic segmentation of urban-scale 3D point clouds: A dataset, benchmarks and challenges |
| title_short | Towards semantic segmentation of urban-scale 3D point clouds: A dataset, benchmarks and challenges |
| title_sort | towards semantic segmentation of urban scale 3d point clouds a dataset benchmarks and challenges |
| work_keys_str_mv | AT huq towardssemanticsegmentationofurbanscale3dpointcloudsadatasetbenchmarksandchallenges AT yangb towardssemanticsegmentationofurbanscale3dpointcloudsadatasetbenchmarksandchallenges AT khalids towardssemanticsegmentationofurbanscale3dpointcloudsadatasetbenchmarksandchallenges AT xiaow towardssemanticsegmentationofurbanscale3dpointcloudsadatasetbenchmarksandchallenges AT trigonin towardssemanticsegmentationofurbanscale3dpointcloudsadatasetbenchmarksandchallenges AT markhama towardssemanticsegmentationofurbanscale3dpointcloudsadatasetbenchmarksandchallenges |