Tracking and Analysis of Pedestrian’s Behavior in Public Places

Crowd management becomes a global concern due to increased population in urban areas. Better of pedestrians leads improved use public places. Behavior pedestrian’s is major factor crowd There are multiple applications available this area but the challenge open complexity and depends on environment. In paper, we have proposed new method for behavior detection. Kalman filter has been used detect using movement based approach. Next, performed occlusion detection removal region shrinking isolate occluded humans. Human verification each human silhouette wavelet analysis particle gradient motion extracted silhouettes. Gray Wolf Optimizer (GWO) utilized optimize feature set then classification Extreme Gradient (XG) Boost classifier. Performance evaluated data from avenue UBI-Fight datasets, where both different The mean achieved accuracies 91.3% 85.14% over Avenue respectively. These results more accurate as compared other existing methods.