Abstract
Algorithms of the visual odometry for intelligent vehicles are reviewed. Firstly, the basic principle of the visual odometry is introduced. Secondly, the classification and comparison are made according to types of visual information, the configuration of cameras and the calculation framework. Thirdly, the key techniques including feature selection and matching, robust motion estimation in two-frame visual odometry system are analyzed. Finally the development trends of the visual odometry algorithms are discussed.
| Original language | English |
|---|---|
| Pages (from-to) | 214-220 |
| Number of pages | 7 |
| Journal | Binggong Xuebao/Acta Armamentarii |
| Volume | 33 |
| Issue number | 2 |
| Publication status | Published - Feb 2012 |
Keywords
- Information processing
- Intelligent vehicles
- Motion estimation
- Visual odometry