Research Article
Online Hierarchical Sparse Representation of Multifeature for Robust Object Tracking
Table 2
The average position errors (pixels).
| Video | IVT | L1 | MIL | OAB | VTD | Our method |
| Car4 | 16.31 | 209.22 | 55.10 | 103.49 | 34.46 | 2.97 | CarScale | 42.64 | 53.52 | 43.63 | 15.66 | 29.35 | 8.44 | CarDark | 22.03 | 19.85 | 48.41 | 3.90 | 23.29 | 1.92 | FaceOcc2 | 16.21 | 15.54 | 18.17 | 24.32 | 10.69 | 8.42 | Freeman1 | 6.85 | 61.74 | 18.98 | 25.44 | 10.71 | 7.44 | Girl | 29.37 | 3.76 | 14.11 | 8.55 | 11.70 | 3.63 | Shaking | 87.90 | 21.05 | 9.59 | 144.66 | 13.42 | 7.11 | Woman | 186.70 | 151.79 | 128.08 | 63.76 | 119.01 | 3.27 | Jogging | 128.29 | 148.81 | 131.14 | 24.29 | 121.31 | 6.54 |
| Average | 59.58 | 76.14 | 51.91 | 46.01 | 41.54 | 5.53 |
|
|
Note: the optimal result is shown as bold and the suboptimal one as italic.
|