Abstract | ||
---|---|---|
This paper presents a novel and efficient object tracking method based on multi-sensor information fusion. Based on the popular detection-tracking framework, we consider the tracking process as 3 conditions and the fusion strategy can be adjusted in different conditions adaptively by designing a new online positive and negative sample classifier selecting method with the guidance of the depth information from the laser scanner. The results of our experiments show good robustness and performance when facing extreme cases such as the object rotation, long-period occlusion, and high similarity between object and background. |
Year | DOI | Venue |
---|---|---|
2012 | 10.1145/2382336.2382361 | ICIMCS |
Keywords | Field | DocType |
depth information,tracking process,fusion strategy,extreme case,different conditions adaptively,real-time pedestrian learning-tracking,high similarity,good robustness,object rotation,multi-sensor information fusion,efficient object tracking method,data fusion | Computer vision,Pedestrian,Laser scanning,Pattern recognition,Computer science,Fusion,Robustness (computer science),Sensor fusion,Video tracking,Artificial intelligence,Classifier (linguistics),Information fusion | Conference |
Citations | PageRank | References |
0 | 0.34 | 8 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Shan Gao | 1 | 88 | 15.15 |
Zhenjun Han | 2 | 176 | 16.40 |
Yang Xu | 3 | 0 | 0.34 |
Qixiang Ye | 4 | 913 | 64.51 |
Jianbin Jiao | 5 | 367 | 32.61 |