Foreign Objects Removal for Self-driving Mapping
| Autor: | Ming-Chun Tseng, Fay Huang, Jheng-Lun Liou, Pu-Sheng Hu, Augustine Tsai |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
0209 industrial biotechnology
Computer science business.industry Deep learning Frame (networking) Point cloud 02 engineering and technology 020901 industrial engineering & automation Self driving 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Computer vision Artificial intelligence business |
| Zdroj: | ICCE-TW |
| DOI: | 10.1109/icce-taiwan49838.2020.9258324 |
| Popis: | A clean map free of foreign objects is essential for self-driving car navigation without compromising localization accuracy. An idea map scanning is to start with a static environment, but it is less likely to prevent cars, cyclists and pedestrians from entering the scene. The situation can result in unwanted noises in the final maps. In this paper, we propose an end-to-end deep learning method to mitigate the problems. The foreign objects are detected and removed in each raw point cloud frame before map generation. Qualitative experiments show the efficacy of the method. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|