Abstract
Autonomous ski robots require robust detection of ski gates when using sensors. Various sensors including 3D LiDAR, 2D LiDAR, stereo systems and cameras might have the potential for detecting ski gates. In this paper, we consider a camera as the main sensor for detecting ski gates by considering processing time and cost. However, cameras have difficulties in coping with variations due to illumination variations. We present a robust algorithm for detecting ski gate using color information. Color differences are used for detecting red and blue ski gates. Binarization using color differences is first performed then morphological erosion and dilation is applied to remove small background objects. Finally, a connected component analysis is performed to find the ski gates. Experimental results using images acquired from real ski slopes demonstrate the feasibility of the presented algorithm.
| Translated title of the contribution | Detection of ski gates using color information |
|---|---|
| Original language | Korean |
| Pages (from-to) | 1123-1127 |
| Number of pages | 5 |
| Journal | Journal of Institute of Control, Robotics and Systems |
| Volume | 24 |
| Issue number | 12 |
| DOIs | |
| State | Published - 2018 |
Keywords
- Autonomous navigation
- Robot vision
- Ski gates detection
- Skiing robot