When uploading point cloud data into Prevu3D, it is important to understand the different types of files and metadata that can be provided. Prevu3D supports a wide range of formats produced by terrestrial laser scanners, SLAM systems, and other lidar devices.
Although Prevu3D can visualize point cloud data directly, certain features such as shading, slicing, and meshing require additional information, such as normals or scan positions. Providing a complete, well structured dataset ensures the best visual and processing results.
Understanding Point Cloud Types
Point clouds come in many forms, and not all contain the information needed for advanced processing. Before uploading, you should be familiar with the basic terminology used in 3D scanning workflows.
Refer to the Data collection article for a comprehensive introduction.
What We Need to Create a Mesh
To produce a high quality mesh from your point cloud, Prevu3D needs the ability to determine the normal (surface direction) for each point. This can come from:
Existing normals stored inside the file
Structured scan data that preserves scanner positions
A trajectory file for SLAM devices
Optional but Recommended: Imagery
If imagery is available, Prevu3D uses the original photos to generate high quality textures and enable features such as photosphere navigation. Image-based texturing produces sharper, more detailed visual results.
If no imagery is provided, Prevu3D will generate the texture using the RGB values stored in the point cloud. This produces a correct but less detailed and slightly blurrier texture compared to one generated from original photos.
Important:
Even without images, the geometry quality of the resulting mesh remains the same. The difference affects only the visual appearance, not the underlying 3D accuracy.
Terrestrial Laser Scanning Data
Prevu3D supports the meshing of high resolution terrestrial scanners using the vendor neutral E57 file format.
The E57 file format is a compact, vendor-neutral format for storing point clouds, images, and metadata produced by the laser scanners. The supported E57 is either structured (setup locations included) point cloud with color imagery, or unified (merged) point cloud with color imagery and normal information
With usual terrestrial datasets, we compute normals using the stations position (hence the need for terrestrial datasets to be structured)
Recommended Export Approach: One E57 Per Station
While some scanners allow you to export a single, merged E57 file that contains all stations, we strongly recommend exporting one E57 file per station instead of one large file.
This approach offers several advantages:
Easier troubleshooting: If one station turns out to be corrupted, noisy, or misaligned, you only need to re-export or re-upload that single file instead of regenerating the entire dataset.
More efficient processing: Processing multiple smaller E57s is often faster than processing one very large file.
More predictable uploads: Smaller files reduce the risk of upload interruptions or timeouts.
Both approaches are supported, but exporting one E57 per station provides the most flexibility and is the preferred method for long-term dataset management.
Info
Make sure to look at the Supported Devices list for more details about the process to get the required & optimal files. If you don’t see your device in the list, please contact us.
SLAM Scanning Data
Prevu3D supports the meshing of SLAM scanners.
With SLAM datasets there are multiple cases:
The pointcloud already contains a normal for each point, so we don’t need to compute any additional information.
Other devices produce both a point cloud and a trajectory. If the format is correct, we can use the trajectory to compute a normal for each point.
Considering there’s no official standard yet in regards of SLAM data, we have a device based upload approach. It will ask you the required files.
Pointcloud (Typically e57 or LAS/LAZ)
Trajectory (if your pointcloud doesn’t contains normal)
Picture (can be embedded inside e57 or external)
Picture position (can be embedded inside e57 or external)
Info
Make sure to look at the Supported Devices list for more details about the process to get the required & optimal files. If you don’t see your device in the list, please contact us.
Structured vs. Unstructured vs. Unified pointcloud
It can be a bit tricky to distinguish the difference between Structured, unstructured and Unified pointcloud terminology. We have an interesting article describing the differences between them.
https://www.prevu3d.com/news/differences-between-structured-unified-and-unstructured-scan-data/