Free LiDAR Point Cloud Manual Classification Tools for LAS and LAZ QA

Need a free point cloud classification brush or paint tool for manual LiDAR QA? This guide covers the free desktop workflows teams use to hand-label LAS and LAZ files, plus when it is faster to switch from brush-based editing to automated classification.

Manual point cloud classification involves selecting and labeling groups of points by hand using classification tools. While time-intensive, manual classification provides precise control over how points are categorized, making it essential for:

• Quality control — Verifying automated results
• Training data creation — Building labeled datasets for ML
• Edge cases — Handling what algorithms miss

OPEN SOURCE

Best for detailed manual editing, visual inspection, and learning

The most popular free point cloud viewer with manual classification capabilities. Supports LAS, LAZ, E57, and many other formats.

• Powerful selection tools & handles large datasets
• Extensive format support (LAS, LAZ, E57)
• Free for personal and commercial use (GPL)

// CloudCompare Workflow

1. Load point cloud (File → Open)

2. Use segmentation tool (scissors icon)

3. Right-click → Edit → Scalar Fields

4. Set SF Value to ASPRS class code

5. Export as LAS

// LAStools lasview Workflow

$ lasview -i yourfile.laz

Press ‘e’ to enter edit mode

Drag to select points

Press 2-9 for classification

Ctrl+S to save

FREE FOR NON-COMMERCIAL

Best for quick visual inspection and spot classification

A lightweight LAS/LAZ viewer with basic classification editing. Faster than CloudCompare for simple tasks.

• Fast, lightweight, keyboard-driven
• Quick spot-fixes and verification

Commercial use requires license

Let AI do 90% of the work, manually verify the rest

Why spend hours manually classifying every point when AI can handle most of it? Upload your LAS/LAZ, get automatic classification into 12+ ASPRS classes in minutes, then manually verify or adjust in CloudCompare.

• 50 hectares free — no credit card
• Minutes instead of hours
• No software to install

Polygon Selection: Draw boundaries around point groups. Best for large, distinct features like buildings.

Brush Selection: Paint-style selection for organic shapes. Works well for vegetation boundaries.

Height Filtering: Select points within elevation ranges. Useful for isolating ground points.

Cross-Section View: Slice through the cloud to see vertical profiles. Essential for ground vs vegetation.

01 Start with ground classification — it defines the foundation

02 Work in sections — divide large datasets into manageable tiles

03 Use multiple views — top-down, profile, and 3D perspective

04 Color by classification — visualize progress and catch errors

05 Save frequently — don’t lose hours of work

Manual Classification

Small datasets (under 10M points)

Training data creation for ML

Quality control verification

Complex edge cases

Budget constraints (time > money)

Automated (AI) Classification

Large datasets (10M+ points)

Production workflows

Repetitive classification tasks

Time-critical projects

Consistent, reproducible results

Best approach: Use AI for the initial 90%, then manually refine problem areas. This hybrid workflow combines automation speed with manual precision.