Abstract | ||
---|---|---|
We developed a poultry tracking system for analyzing the behavior of poultry infected with avian influenza using a camera. The trackers employed in our system robustly track poultry that make contact and occlude each other in a narrow isolator during infection experiment using a particle filtering algorithm. This system has two kinds of trackers: poultry trackers and the exploring trackers. We used one poultry tracker for each target in the isolator. An exploring tracker searches and corrects a poultry tracker that failed. Further, complete overlap among trackers is avoided marking tracker positions on an image. We evaluated this system using 5 min video data on 10 healthy poultry. The results showed that our system can recognize and track poultry effectively; however, the durations for which poultry were tracked were not sufficient. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1109/ROBIO.2009.4913289 | 2008 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND BIOMIMETICS, VOLS 1-4 |
Keywords | Field | DocType |
Image processing, Particle filter, Tracking multiple targets | BitTorrent tracker,Computer vision,Control theory,Particle filter,Image processing,Filter (signal processing),Tracking system,Robustness (computer science),Artificial intelligence,Particle filtering algorithm,Engineering,Isolator | Conference |
Citations | PageRank | References |
0 | 0.34 | 3 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Toshiyuki Fujii | 1 | 0 | 1.69 |
Hiroshi Yokoi | 2 | 383 | 92.58 |
Tatsuya Tada | 3 | 0 | 0.34 |
Kotaro Suzuki | 4 | 0 | 0.34 |
Kenji Tsukamoto | 5 | 1 | 0.70 |