Table 5 Correlation coefficient between pressure measured at the doorjamb and different crowd measures based on tracked trajectories. For quantities based on pedestrian positions an area having 75 cm of radius and centered at the exit (coordinate (0, 0)) was taken to compute the average for each condition. Note: since pressure sensors stopped for some trials or synchronization was not possible with the videos, the correlation coefficient has been computed based on each condition (considering only competitive experiments) and not for each trial as in previous analyses. Thus, in this case only four values (corresponding to the competitive cases in Fig. 18) are used in computing the correlations and this can also explain the very high p-values obtained in this analysis.
Quantity | Unit | Region considered | Correlation coefficient |
|---|---|---|---|
Crowd angle | [rad] | Exit location (door) | 0.875 (p = 0.125) |
Mean flow rate | [\({\hbox {s}}^{-1}\)] | Exit location (door) | 0.665 (p = 0.335) |
Speed | [m/s] | Exit front area | −0.207 (p = 0.793) |
Density | [\({\hbox {m}}^{-2}\)] | Exit front area | −0.646 (p = 0.354) |
Congestion level | [\({\hbox {m}}^{-1}\)] | Exit front area | 0.836 (p = 0.164) |
Crowd danger | [\({\hbox {m}}^{-3}\)] | Exit front area | −0.051 (p = 0.949) |
