Research Article
Human Action Recognition Based on Fusion Features Extraction of Adaptive Background Subtraction and Optical Flow Model
Table 1
The results of human action recognition on Weizmann database.
| The action categories | Total number of samples | The number of the times of correct recognition | The correct recognition rate (%) |
| Bending | 200 | 200 | 100.0 | Running | 200 | 190 | 95.0 | Walking | 200 | 199 | 99.5 | Jack | 196 | 196 | 100.0 | Jump | 195 | 180 | 92.3 | pjump | 192 | 192 | 100.0 | Sideways | 186 | 186 | 100.0 | Wave2 | 192 | 192 | 100.0 | Wave1 | 180 | 180 | 100.0 | The average recognition rate | 98.5 |
|
|