An Efficient and General Framework for Aerial Point Cloud Classification in Urban Scenarios
| Autor: | Fabio Remondino, E. Özdemir, Alessandro Golkar |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
010504 meteorology & atmospheric sciences
Computer science Generalization Science 0211 other engineering and technologies Point cloud ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Inference 02 engineering and technology Machine learning computer.software_genre 01 natural sciences 021101 geological & geomatics engineering 0105 earth and related environmental sciences aerial point cloud business.industry Deep learning deep learning Power (physics) Photogrammetry Lidar machine learning classification AI General Earth and Planetary Sciences Artificial intelligence State (computer science) business computer |
| Zdroj: | Remote Sensing Volume 13 Issue 10 Pages: 1985 Remote Sensing, Vol 13, Iss 1985, p 1985 (2021) |
| ISSN: | 2072-4292 |
| DOI: | 10.3390/rs13101985 |
| Popis: | 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. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|