An Efficient and General Framework for Aerial Point Cloud Classification in Urban Scenarios
With recent advances in technologies, deep learning is being applied more and more to different tasks. In particular, point cloud processing and classification have been studied for a while now, with various methods developed. Some of the available classification approaches are based on specific dat...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-05-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/10/1985 |
| _version_ | 1797533512914436096 |
|---|---|
| author | Emre Özdemir Fabio Remondino Alessandro Golkar |
| author_facet | Emre Özdemir Fabio Remondino Alessandro Golkar |
| author_sort | Emre Özdemir |
| collection | DOAJ |
| description | With recent advances in technologies, deep learning is being applied more and more to different tasks. In particular, point cloud processing and classification have been studied for a while now, with various methods developed. Some of the available classification approaches are based on specific data source, like LiDAR, while others are focused on specific scenarios, like indoor. A general major issue is the computational efficiency (in terms of power consumption, memory requirement, and training/inference time). In this study, we propose an efficient framework (named TONIC) that can work with any kind of aerial data source (LiDAR or photogrammetry) and does not require high computational power while achieving accuracy on par with the current state of the art methods. We also test our framework for its generalization ability, showing capabilities to learn from one dataset and predict on unseen aerial scenarios. |
| first_indexed | 2024-03-10T11:15:37Z |
| format | Article |
| id | doaj.art-dc73e34a636446b0b38050082b40930f |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T11:15:37Z |
| publishDate | 2021-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-dc73e34a636446b0b38050082b40930f2023-11-21T20:27:45ZengMDPI AGRemote Sensing2072-42922021-05-011310198510.3390/rs13101985An Efficient and General Framework for Aerial Point Cloud Classification in Urban ScenariosEmre Özdemir0Fabio Remondino1Alessandro Golkar2Skolkovo Institute of Technology (Skoltech), 143026 Moscow, Russia3D Optical Metrology (3DOM) Unit, Bruno Kessler Foundation (FBK), 38123 Trento, ItalySkolkovo Institute of Technology (Skoltech), 143026 Moscow, RussiaWith recent advances in technologies, deep learning is being applied more and more to different tasks. In particular, point cloud processing and classification have been studied for a while now, with various methods developed. Some of the available classification approaches are based on specific data source, like LiDAR, while others are focused on specific scenarios, like indoor. A general major issue is the computational efficiency (in terms of power consumption, memory requirement, and training/inference time). In this study, we propose an efficient framework (named TONIC) that can work with any kind of aerial data source (LiDAR or photogrammetry) and does not require high computational power while achieving accuracy on par with the current state of the art methods. We also test our framework for its generalization ability, showing capabilities to learn from one dataset and predict on unseen aerial scenarios.https://www.mdpi.com/2072-4292/13/10/1985aerial point cloudclassificationAImachine learningdeep learning |
| spellingShingle | Emre Özdemir Fabio Remondino Alessandro Golkar An Efficient and General Framework for Aerial Point Cloud Classification in Urban Scenarios Remote Sensing aerial point cloud classification AI machine learning deep learning |
| title | An Efficient and General Framework for Aerial Point Cloud Classification in Urban Scenarios |
| title_full | An Efficient and General Framework for Aerial Point Cloud Classification in Urban Scenarios |
| title_fullStr | An Efficient and General Framework for Aerial Point Cloud Classification in Urban Scenarios |
| title_full_unstemmed | An Efficient and General Framework for Aerial Point Cloud Classification in Urban Scenarios |
| title_short | An Efficient and General Framework for Aerial Point Cloud Classification in Urban Scenarios |
| title_sort | efficient and general framework for aerial point cloud classification in urban scenarios |
| topic | aerial point cloud classification AI machine learning deep learning |
| url | https://www.mdpi.com/2072-4292/13/10/1985 |
| work_keys_str_mv | AT emreozdemir anefficientandgeneralframeworkforaerialpointcloudclassificationinurbanscenarios AT fabioremondino anefficientandgeneralframeworkforaerialpointcloudclassificationinurbanscenarios AT alessandrogolkar anefficientandgeneralframeworkforaerialpointcloudclassificationinurbanscenarios AT emreozdemir efficientandgeneralframeworkforaerialpointcloudclassificationinurbanscenarios AT fabioremondino efficientandgeneralframeworkforaerialpointcloudclassificationinurbanscenarios AT alessandrogolkar efficientandgeneralframeworkforaerialpointcloudclassificationinurbanscenarios |