Turn airborne LiDAR into operational land-cover and forest intelligence.
A reproducible pipeline for point-wise semantic segmentation, GIS-ready outputs and forest analytics, with strong internal performance and demonstrated improvement on a separate geographic holdout.
Airborne LiDAR
XYZ, RGB/CIR/NIR, intensity and reference classifications.
Label-free local geometry
Density, height and surface context at several spatial scales.
Robust segmentation
Supervised training with robust normalization and class balancing.
GIS and forest outputs
Maps, LAZ, grids, indicators, metrics and review priorities.
Manual LiDAR interpretation does not scale.
Large surveys contain millions or billions of points. Teams need consistent maps and a prioritized review process, not another raw data archive.
Slow manual mapping
Reviewing point clouds by hand consumes specialist time and delays projects.
Difficult boundaries
Low vegetation, medium vegetation, roofs and tree crowns require local 3D context.
Underused data
Organizations often own LiDAR but lack a repeatable pipeline that converts it into operational layers.
A practical 3D workflow from raw points to decisions.
Airborne LiDAR, color, infrared and intensity.
Clean, aligned and reproducible train/validation/test data.
Label-free geometry at 2.5 m, 5 m and 10 m.
Robust normalization, focal loss and balanced training.
Semantic layers, forest indicators and technical reports.
Two reference models serve different objectives.
Local-context model
Highest internal macro-F1 and mean IoU in the current comparison.
Robust local-context model
20k balanced blocks and robust normalization for stronger geographic transfer.
Strong gains across all six semantic classes.
| Metric | Original baseline | Best internal mIoU run | Delta |
|---|---|---|---|
| Overall accuracy | 83.03% | 86.49% | +3.45 pp |
| Macro-F1 | 74.82% | 82.60% | +7.78 pp |
| Mean IoU | 63.49% | 72.63% | +9.14 pp |
What changed
The model receives local height, density, color and surface context rather than interpreting each point in isolation.
Strong terrain, trees, buildings and water separation.
The main deployment challenge was domain shift.
Spectral shift
Mean NIR changed from 0.5050 internally to 0.2955 on the external development set.
Terrain shift
Mean block height range changed from 16.99 to 61.58, with a much larger external tail.
Effective correction
Robust coordinate/spectral normalization plus label-free geometry and 20k balanced blocks.
Robustness improved on a separate 32-tile campaign.
| Model | OA | Macro-F1 | mIoU |
|---|---|---|---|
| Strong TW baseline | 72.60% | 63.99% | 49.45% |
| Robust local-context | 80.83% | 73.42% | 60.41% |
Mean IoU over the strong baseline
Overall accuracy improved by 8.23 pp and macro-F1 by 9.43 pp, with 6/6 class IoUs improved.
Improvement carries to unseen, non-overlapping tiles.
| Model | OA | Macro-F1 | mIoU | mIoU excl. water |
|---|---|---|---|---|
| Strong TW baseline | 77.33% | 51.38% | 40.62% | 48.33% |
| Robust local-context | 83.27% | 61.80% | 50.98% | 60.62% |
External mean IoU
OA improved by 5.94 pp and macro-F1 by 10.42 pp. All six class IoUs improved against the strong baseline.
The largest transfer gain appears in buildings.
| Class | Strong baseline IoU | Robust IoU | Delta |
|---|---|---|---|
| Ground | 73.43% | 79.97% | +6.54 pp |
| Low vegetation | 21.23% | 27.70% | +6.47 pp |
| Medium vegetation | 47.28% | 56.46% | +9.18 pp |
| High vegetation | 81.93% | 88.47% | +6.54 pp |
| Building | 17.80% | 50.50% | +32.70 pp |
| Water* | 2.09% | 2.77% | +0.68 pp |
Segmentation becomes practical forest information.
Vegetation ratios
Low, medium, high and combined vegetation by tile or grid cell.
Canopy cover proxy
High and medium-high vegetation ratios from classified points.
Height proxy
Vegetation z95 relative to local ground, or a conservative fallback.
Priority flags
Canopy gaps, sparse cells and unusual vegetation/building patterns.
GIS-native deliverables and a low-risk deployment path.
Operational outputs
Classified LAZ, semantic maps, grids, forest layers and review-priority maps.
Evidence and QA
Global metrics, class metrics, confusion matrices, error analysis and reproducible reports.
Integration
Delivery to QGIS, ArcGIS, PostGIS, GeoServer, web viewers or customer APIs.
1. Discover
Choose one area and one measurable operational problem.
2. Pilot
Process historical data and validate representative zones.
3. Scale
Run tiled inference over a larger territory.
4. Operate
Monitor quality, update models and add future analytics.
A commercially credible LiDAR intelligence MVP with demonstrated geographic transfer.
The system combines label-free local geometry, robust normalization and supervised learning to produce GIS-ready land-cover and forest intelligence. It preserves strong internal quality and improves a strong baseline on a fresh, tile-disjoint external holdout.