By combining a sheer size of 16 billion points with a very uniform spread and an incredible 0.5mm density of the scan, Virtual Wonders managed to produce one of the largest and certainly the densest datasets we’ve seen to date. Their data proves that raw 3D scanned assets are capable of resolving photo-realistic levels of detail on a very large scale.
Visualizing the Paroo Station dataset in real-time was one of our most challenging projects. But it was also an incredible opportunity to push the technology farther than we thought was possible.
City-sized point cloud dataset:
All running in real-time, powered by Unreal Engine 4
An example of how LiDAR point cloud data can work as a base for setting up a virtual scene.
Here, we are combining it with a collection of Speed Trees, Unreal Engine-powered dynamic lighting and shadows, as well as a number of post-processing effects.
The project has been designed to allow free exploration using VR goggles.