This project provides a real-time 3D visualization of real-world multi-pedestrian trajectories, transforming 2D tracking data into an interactive 3D representation. Unlike static plots, it uses PyQtGraph’s OpenGL module to render pedestrians as customizable humanoid models, with individually modeled body parts—head, torso, arms, and legs—allowing realistic depiction of posture and motion.
The visualization incorporates environmental geometry, including walls, exits, and pillars, to accurately reflect the spatial context of the original environment. Pedestrians move through this 3D scene in real-time according to recorded trajectories, enabling interactive exploration such as rotating the viewpoint, zooming, and inspecting individual agents.
This setup supports detailed analysis of crowd behavior, space usage, and human-environment interactions without relying on predictive simulation models, providing a clear and immersive way to study pedestrian movement in existing environments.
Figure: Sample 3D simulation of pedestrians with custom humanoid models in an environment.
D. Brscic, T. Kanda, T. Ikeda, T. Miyashita, "Person position and body direction tracking in large public spaces using 3D range sensors", IEEE Transactions on Human-Machine Systems, Vol. 43, No. 6, pp. 522-534, 2013
✔️ Customized 3D humanoid agents created (body, head, arms, legs)
✔️ Real-time trajectory playback using PyQtGraph OpenGL 3D
🔄 Implementing environment import, collision detection, and user interaction controls
📈 Planning to integrate statistical analysis tools