ICFICE 2019
Current Result Document :
| 한글제목(Korean Title) |
Optical Flow-based Unmanned Aerial Vehicles Tracking |
| 영문제목(English Title) |
Optical Flow-based Unmanned Aerial Vehicles Tracking |
| 저자(Author) |
Sohee Son
Jinwoo Jeon
Injae Lee
Haechul Choi
|
| 원문수록처(Citation) |
VOL 11 NO. 01 PP. 0295 ~ 0297 (2019. 06) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
Tracking for unmanned aerial vehicles (UAVs) in natural scene sequences is a challenging task due to the small size of the unrigid targets and the large inter-frame disparity by fast motion. To deal with the small size and the large disparity, a point-based approach would be better than block-based and template-based ones. The proposed method measures motion flows of feature points by means of the optical flow and tracks UAVs using only reliable motion flows. In addition, feature points are updated in the way that feature points with unreliable motion flows are rejected to reduce mistracking and new feature points are periodically added to keep a sufficient number of feature points. Experimental results show that UAVs with small size are well tracked from natural scene sequences.
|
| 키워드(Keyword) |
Computer Vision
Object Tracking
Optical Flow
UAV
|
| 파일첨부 |
PDF 다운로드
|
