Table 3.
List of sequences of images used as ground truth. These images present complex scenarios. The first two sequences represent scenarios with shadows caused by different light sources and both color and texture of skin in some frames, that are too similar to background scenario. The third one, represents a scenario with high noise degree, the reflections are caused by the windows and the moving objects are represented by small zones.
| No. | Place | Source | Num. Img.1 | bf Num. Train. 2 |
|---|---|---|---|---|
| 1 | Train Station | PETS 2007 | 300 | 100 |
| 2 | Train Station | PETS 2007 | 300 | 100 |
| 3 | Outdoors Park | PETS 2001 | 500 | 100 |
1
Total number of images used.
2
Number of images used to train background model approach.