01
Upload
Drag and drop your LAS or LAZ file. No installation, no setup.
AI classification
Classification That WorksOn the First Try
No parameter tuning. No steep-terrain headaches. Upload your LAS file and get 12+ classes automatically: ground, vegetation, buildings, water, wires and more.
No credit card required • 50 Ha of free classification
12+ classes
Standard ASPRS codes
Standard ASPRS codes
No tuning
Zero parameters
Zero parameters
Any terrain
Flat or mountain slope
Flat or mountain slope
Manual editing
Brush, permute, filter
Brush, permute, filter
How it works
From raw LAS to classified cloud in three steps
02
Classify
The AI analyses the cloud and classifies it into 12+ categories. No parameters to tune.
03
Refine & export
Fix edge cases with brush selection, class permutation and filtering, then export the cloud, DTM, DSM or contours.
Coverage
12+ classification categories
Standard LAS classes, assigned with a single click.
| Group | Classes | Typical use |
|---|---|---|
| Terrain | Ground | DTM, contours, volumes |
| Vegetation | Low · Medium · High, Hedge | Forestry, encroachment, CHM |
| Structures | Building, Roof object, Bridge deck | Footprints, urban planning |
| Utilities | Wire, Tower, Pole | Corridor analysis, clearance |
| Surfaces | Water | Hydrography, shoreline |
| Other | Vehicle, Fence / Wall | Site inventory, site context |
Where it earns its keep
Classification power for key industries
Terrain
Ground Classification on Any Terrain
Flat areas and mountain slopes, equally well, out of the box.
The AI separates ground from vegetation, buildings and noise without manual adjustment. Traditional software needs careful parameter tuning on steep terrain; this does not.
- No slope thresholds or filtering parameters to set
- Trained on diverse terrain, not tuned per project
- Cleaner ground means a cleaner DTM downstream
Vegetation
Vegetation Classification
Low, medium and high vegetation, separated automatically.
Built for forestry analysis, vegetation management and environmental monitoring, where the split between storeys is the whole point.
- Low, medium and high vegetation as separate classes
- Feeds canopy height models and forest inventory
- Drives encroachment analysis on utility corridors
Infrastructure
Wires, Towers & Buildings
Dense urban and utility corridors, without hand-tuning.
Buildings and roof objects are classified accurately even in dense urban environments, and the AI distinguishes structures from vegetation and adjacent terrain. Wires, towers and poles come out cleanly for corridor analysis.
- Buildings and roof objects in dense urban scenes
- Wires, towers and poles for utility corridors
- Water bodies identified, noise from birds and artefacts filtered
Customer stories
Trusted by survey professionals worldwide
The powerful classification is the best feature. It saves a lot of time and money compared to manual classification or expensive solutions.
Emilio SegoviaHead of Operations, PDM 4 Cero
We saved a lot of time classifying our point clouds. Not only classified data, but also DEM, DSM, DXF exports we could use directly in our analysis software.
Marcel AertssenSurvey Analyst, SkySense UAV Solutions
Excellent software. Processing is extremely fast and highly accurate. Exemplary DSM and DTM outputs.
Theoharis PapadiamantisSurveying Engineer, Geodimetro
Lidarvisor has transformed my LiDAR post-processing workflow. The speed of results is impressive, all the useful data is correlated, and what I find exceptional is the DXF export function.
Carolina CollaroArchitect | University of Jaén, Spain
We used to outsource our LiDAR processing. With Lidarvisor, we brought it in-house and cut our costs significantly while keeping full control over the workflow.
Bodie Lloyd KiroriManaging Director | Geolens Engineering, Papua New Guinea
We tested Lidarvisor and achieved excellent results using a single platform. Previously, we needed up to three different programs to get similar results, which took us far more time and effort.
Eduardo BonillaProject Manager | Topcom S.A. de C.V., El Salvador
Questions
Frequently asked questions
How accurate is the AI classification?
Lidarvisor uses deep learning models trained on diverse aerial LiDAR datasets, and achieves high accuracy across terrain types from flat urban areas to steep forested slopes. It handles edge cases that trip up rule-based systems, and results are reliable on the first try without manual adjustment.
Can I manually edit classifications?
Yes. Brush selection picks points, class permutation reassigns them to any class, and filtering isolates a class for focused editing. These tools are built in, so fixing edge cases after AI classification never means exporting to another application.
Does it work on steep terrain?
Yes. Unlike traditional classification software that needs careful parameter tuning for steep slopes, the model was trained on diverse terrain and handles it automatically. There are no slope thresholds or filtering parameters to adjust.
What file formats are supported?
Upload point clouds in LAS, LAZ (compressed LAS) or COPC, the industry-standard formats. Export classified results as LAS, or generate DTM and DSM rasters and contours in DXF, SHP or GeoJSON.
How long does classification take?
It depends on file size and point density, but most datasets classify within 5 to 15 minutes. Large projects can take longer, and you get an email when processing is complete.
Do I need to install any software?
No. Lidarvisor is entirely web-based. Open your browser, upload your file and start processing. No plugins, no installation, no configuration, on any modern browser.
Documentation
Learn classification
How to classify your LiDAR point clouds with Lidarvisor.
Ready to simplify your classification workflow?
Start with 50 Ha of free classification. No credit card required. See results in minutes.
Create free accountQuestions? Contact our team or check our pricing plans