Advanced
Managing LiDAR point clouds
Updated Apr 2026·8 min read
LiDAR point clouds are the heaviest and most valuable data. Neobora manages them in the cloud: optimized upload, smooth visualization by level of detail, symbology by classification and error editing.
Visualization by level of detail
When you upload a LAS/LAZ cloud, Neobora builds a multi-resolution index. This lets you explore hundreds of millions of points in the browser without downloading them all: more points load as you zoom in.
i
Performance. You don’t need to split your flights manually. Upload the full project; the indexing takes care of serving only what’s visible at each moment.
Symbology and classification
You can color the cloud by different attributes depending on what you need to analyze:
- By classification: ground, vegetation (low/medium/high), buildings, conductors, towers…
- By elevation: color ramp by height (Z).
- By intensity or RGB if the flight includes it.
Editing and fixing classes
AI automatic classification handles most of the work, but quality control remains human. From the editor you can:
- 1Select points with polygon, brush or range tools.
- 2Reassign the class of the selection (e.g. noise → unclassified).
- 3Isolate a class to review it cleanly.
- 4Save: changes version the layer without losing the original.
i
Recommended workflow. Run the AI classification first, review the low-confidence zones the model flags and correct only those. It’s the fastest way to guarantee quality without editing point by point.
Performance and best practices
- Work with the cloud in the cloud: avoid downloading the LAZ to edit locally.
- Use folders by zone or flight date to compare changes over time.
- Publish a private geoportal so technicians can review without touching the source data.
Was this article helpful?