- Parks are large with many people, vehicles and facilities across separate systems.
- Video twin = live video + 3D model, giving one spatial view for dispatch.
- Existing cameras and 3D assets are reused.
1. The Park Management Problem
An industrial or business park spans roads, buildings, greenery and boundaries, with cameras and management systems that are separate. Operators lose the spatial relationship between feeds — a person or vehicle moving across zones is hard to follow, and incidents are hard to locate on one screen.
2. The Solution: Video Twin (Video + 3D)
Video twin overlays live video onto a 3D park model at exact spatial positions. Operators see the real site inside the model — a vehicle driving through the gate appears in the 3D scene at the correct location. This is SpatioForge, fusing video with 3D for Unreal, Unity, Cesium and Three.js.
3. What It Covers
- Roads and boundaries — stitched into one continuous view, eliminating blind spots.
- People and vehicles — tracked across camera boundaries without switching screens.
- Facilities and buildings — mapped to the 3D model for one-screen dispatch.
4. Performance Numbers
| Metric | Value |
|---|---|
| Max channels (single server) | 156 |
| End-to-end latency | Within 200 ms |
| 3D engines | Unreal / Unity / Cesium / Three.js |
| Camera reuse | RTSP / ONVIF / GB28181 |
5. FAQ
Can existing 3D models be reused?
Yes. Fusion does not replace existing 3D assets; the SDK adapts to the engine version while models and business logic stay unchanged.
Can existing cameras be reused?
Yes. Any camera exposing RTSP, ONVIF or GB28181 can join the pipeline, keeping the park's front-end investment.
How many video channels can one scene take?
No fixed ceiling — concurrent channels scale with hardware and scene complexity; judge by channels shown live on screen.