Noise & outliers
Low points, atmospheric interference and sensor artefacts contaminate the data.
Point cloud preprocessing
Clean and OptimizeBefore You Process
Raw LiDAR is messy. Noise, excessive density and missing colour make analysis harder than it needs to be. Lidarvisor cleans and optimises your point clouds automatically, so every downstream step starts from better data.
No credit card required • 50 Ha of free processing
Smoothing
Noise reduction
Noise reduction
Thinning
Density optimization
Density optimization
Colorization
RGB from satellite
RGB from satellite
Noise removal
Outlier detection
Outlier detection
The problem
What raw LiDAR arrives with
Traditional software answers these with hours of manual cleanup and parameter tweaking. Lidarvisor automates the whole preprocessing pass.
Excessive density
Modern sensors capture millions of points per second, making files too large to process efficiently.
No colour information
Monochrome point clouds are hard to interpret and hard to present to a client.
Surface irregularities
Minor variations produce rough terrain models that do not reflect reality.
What it does
Five passes, applied in the right order
Combine any of them in a single workflow — Lidarvisor sequences them for you.
Surface smoothing
Point Cloud Smoothing
Reduce noise while preserving real terrain features.
Surface-aware smoothing with quadratic surface fitting smooths the cloud without destroying the detail that matters.
- Three levels: soft, medium or strong
- Terrain-preserving — adapts to local surface geometry
- Cleaner DTMs from smoothed ground points
Density optimization
Point Cloud Thinning
Reduce file size without losing data quality.
Grid-based subsampling removes redundant points while keeping the structure and accuracy of the cloud. Built for dense UAV and terrestrial scans.
- Configurable grid size in centimetres
- Faster processing from smaller files
- Accuracy maintained through strategic point selection
Satellite imagery
Automatic Colorization
Turn a monochrome cloud into a full-colour 3D scene.
RGB colour is applied from satellite imagery automatically, making the data easier to interpret and far easier to present. No manual alignment.
- Colours sourced from Azure Maps satellite imagery
- Automatic georeferencing — no manual alignment
- Colour preserved in LAZ, LAS and other exports
Outlier detection
Noise Classification
Separate noise from valid data automatically.
Low points and high noise are detected and classified separately, so your analysis only ever runs on clean data.
- Low point detection — finds points below ground
- High noise separation — isolates atmospheric artefacts
- Download with noise filtered out, or kept and classified
Ground control points
Control Point Alignment
Anchor the cloud to your ground survey.
Import surveyed control points and Lidarvisor measures the cloud against them — every point checked individually, with its offset visible in the viewer. When a correction is worth applying, it is applied during processing.
- Reads the point files your GNSS and survey packages export
- Free automatic accuracy check against every control point
- Correction options from a simple vertical shift to a full local fit
- The delivered report states the accuracy that was measured, not assumed
How it works
Preprocessing runs inside your normal workflow
01
Upload
LAS, LAZ or any other supported format.
02
Configure
Pick smoothing level, thinning grid and colorization.
03
Process
Lidarvisor runs it automatically — no parameters to tune.
04
Download
Export clean, optimised data ready for analysis.
Who it's for
Built for your workflow
Surveyors
Smooth ground points for accurate terrain models.
UAV Operators
Thin dense scans for faster processing.
Client Deliverables
Colourized point clouds a client can actually read.
Quality Assurance
Remove noise before analysis so results hold up.
Questions
Frequently asked questions
Does preprocessing affect my original data?
No. Lidarvisor always preserves your original point cloud. Preprocessing creates a new, optimised version while keeping the original intact, and you can download either at any time.
How does smoothing preserve terrain features?
It uses quadratic surface fitting that adapts to local geometry. The algorithm smooths noise while respecting terrain breaks, edges and real surface variation, and displacement limits prevent over-smoothing.
What grid sizes are available for thinning?
You configure the grid size in centimetres. Common values run from 5 to 50 cm depending on your accuracy requirements and target file size. Smaller grids retain more detail; larger grids produce smaller files.
Where does the colorization imagery come from?
Satellite imagery from Azure Maps. The system georeferences your point cloud coordinates against the imagery and applies RGB values per point. No manual alignment is needed.
Can I apply multiple preprocessing steps?
Yes. Smoothing, thinning, colorization and noise removal can be combined in a single workflow, and Lidarvisor applies them in the optimal order automatically.
Can Lidarvisor align my point cloud to ground control points?
Yes. Import the control points from your ground survey and Lidarvisor checks the cloud against them automatically. If a correction helps, it recommends one — from a simple vertical shift to a full local fit — and applies it during processing. A correction is only ever applied when it measurably improves the result. The ground control points guide walks through the whole flow, from the file your instrument exports to the accuracy statement in your report.
Does the control point check cost credits?
No. Importing control points and the automatic accuracy check are free. Only the alignment itself runs as part of a normal processing run, and the delivered report then states the accuracy that was actually measured against your points.
Documentation
Learn preprocessing
How to preprocess and clean your point clouds in Lidarvisor.
Ready to clean up your point clouds?
Every Lidarvisor account includes full access to the preprocessing tools. Start with 50 Ha of free processing.
Start freeQuestions? Contact our team or check our pricing plans