Research Article
An Algorithm of Occlusion Detection for the Surveillance Camera
Table 1
The result of recognition with different similarity thresholds.
| Similarity threshold | Accuracy rate (%) | False positives rate (%) | False negatives rate (%) |
| 0.8 | 90.80 | 6.90 | 2.30 | 0.9 | 89.65 | 6.90 | 3.45 | 1.0 | 90.80 | 3.45 | 5.75 | 1.1 | 89.65 | 2.30 | 8.05 | 1.2 | 88.54 | 1.15 | 10.34 |
|
|