Understanding Your Results

After processing, your project tree fills up with new assets — a classified point cloud, terrain models, vector layers, and reports. This page explains what each one is and how to work with it.

The Project Tree After Processing

Once processing completes, the left panel shows an expanded tree of results organized by type. Here is what you might see (depending on which options you enabled):

📁 My Project
├── 📄 Original Point Cloud
├── 📄 Classified Point Cloud
├── 🗺️ DTM (Digital Terrain Model)
├── 🗺️ DTM Hillshade
├── 🗺️ DSM (Digital Surface Model)
├── 🗺️ CHM (Canopy Height Model)
├── 🗺️ Slope Map
├── 🗺️ TIN (Triangulated Irregular Network)
├── 📐 Contour Lines (Major)
├── 📐 Contour Lines (Minor)
├── 📐 Elevation Grid
├── 📐 Main Breaklines
├── 📐 Detailed Breaklines
├── 📐 Exhaustive Breaklines
├── 📐 Building Footprints
├── 📐 Bridges
├── 📐 Roads
├── 📐 Rails
├── 📐 Water Body
├── 📐 Shoreline
├── 📐 Cadastral Parcels
├── 📐 Tree Crowns
├── 📐 Vegetation Areas
├── 📐 Tree Tops
├── 📐 Power Lines
├── 📐 Towers
├── 📐 Encroachment Findings
│   ├── 📐 Buffer Zone 1
│   ├── 📐 Buffer Zone 2
│   ├── 📐 Buffer Zone 3
│   ├── 📐 Clearance Zone
│   ├── 📐 Right-of-Way Zone
│   └── 📐 Fallen Tree Risk
├── 📋 Topographic Map
├── 📋 Forest Inventory
└── 📋 Vegetation Encroachment Report

Not all of these will appear — only the products you selected during processing. A layer that ran but legitimately found nothing (no water on a dry site, no cadastre coverage in your country) appears marked as empty rather than failing.

The Classified Point Cloud

This is the most important result. Every point in your original cloud has been automatically assigned a classification by Lidarvisor's AI. The classification tells you what each point represents in the real world.

Classification Classes

Here are the classes Lidarvisor can identify. The number after each class is its standard classification code — the same code stored in LAS files and shown next to the class name in the app:

Class	Description	Default Color
Never Classified (0)	Points that have never been assigned a class	Salmon
Unclassified (1)	Points the AI could not confidently assign to a class	Salmon
Ground (2)	The bare earth surface — soil, rock, pavement at ground level	Blue
Low Vegetation (3)	Grass, shrubs, and plants below approximately 0.5 meters	Light green
Medium Vegetation (4)	Bushes and small trees between approximately 0.5 and 1.5 meters	Yellow-green
High Vegetation (5)	Trees and tall vegetation above approximately 1.5 meters	Green
Building (6)	Rooftops and building structures	Red
Low Point (7)	Erroneous points below the ground surface	White
Water (9)	Lakes, rivers, ponds, and other water surfaces	Deep blue
Rail (10)	Railway tracks	Teal
Road Surface (11)	Paved and unpaved road surfaces	Brown
Wire — Guard / Shield (13)	Shield and earth wires strung above the conductors	Orange
Wire — Conductor / Phase (14)	The current-carrying power line cables	Cyan
Transmission Tower (15)	Power line towers and pylons	Light gray
Wire-Structure Connector (16)	Insulators and fittings connecting wires to structures	Pink
Bridge Deck (17)	Bridge surfaces	Dark purple
High Noise (18)	Erroneous points well above the actual surface	White
Overhead Structure (19)	Canopies, awnings, and other overhead constructions	Dark teal
Ignored Ground (20)	Ground points excluded from terrain modeling	Gray
Snow (21)	Snow-covered surfaces	Light gray
Temporal Exclusion (22)	Points excluded because they were captured at a different time	Yellow
Roof Object (23)	Chimneys, antennas, dormers, and other objects on roofs	Dark gray
Vehicle (24)	Cars, trucks, and other vehicles	Dark red
Pole (25)	Utility poles, light poles, sign poles	Coral
Fence / Wall (26)	Fences, walls, and barriers	Violet
Hedge (27)	Hedgerows and trimmed vegetation borders	Forest green
Solar Panel (28)	Rooftop or ground-mounted solar panels	Gold

Viewing Classification Results

1. Make sure the Classified Point Cloud is checked (visible) in the project tree.
2. Switch the visualization mode to Classification using the dropdown on the point cloud node.
3. Expand the classified point cloud node to see individual classes.
4. Tick or untick individual classes to isolate what you want to see.

For example:

• Untick everything except Ground to see just the bare earth.
• Untick Ground and vegetation to see just buildings, power lines, and man-made objects.
• Untick noise classes (Low Point, High Noise) — these are usually hidden by default.

Customizing Class Colors

If you want different colors for any class:

1. Click the small colored circle next to the class name in the project tree.
2. A color picker palette appears — choose your preferred color.
3. Click Reset to go back to the default color.

Your custom colors are saved and will be remembered next time you open the project.

How Vector Extraction Refines the Classification

The AI classifies the cloud point by point, and on some surfaces — calm water, bridge decks, track beds — the per-point result is patchy. When you enable the extraction of water, roads, bridges, or rails, Lidarvisor writes what those extractions learned back into the classified point cloud, so the point cloud you download agrees with the vector layers in the same delivery.

The corrections are applied in a deliberate order so the right class wins wherever features overlap:

1. Water first. Points on the detected water surface become Water (9), even where they were originally labeled Ground or left unclassified. Stray water labels away from any detected water — typically wet asphalt or dark roofs the AI mistook for water — are demoted to Unclassified (1). A height ceiling at the local water level keeps overhanging vegetation, bridges, and boats from being swept into the water class.
2. Roads. Ground points inside the detected road surfaces become Road Surface (11).
3. Bridges. Points at deck level inside each bridge polygon become Bridge Deck (17) — vehicles parked on the deck keep their own class. Parapet points just above the deck become Fence / Wall (26), as do the bridge's supports below the deck, while the water flowing underneath stays Water.
4. Rails last. The track corridor becomes Rail (10), reclaiming the track bed even where the road or bridge steps above had claimed it — so a railway crossing a bridge is still Rail. Vegetation, wires, and structures above the track keep their classes.

Two things to know about these corrections:

• They run once, at processing time. If you later edit a vector layer in the viewer, the vector files are updated but the point classification is not re-run.
• The classification statistics and the per-class filters in the viewer reflect the corrected classes.

Terrain Models (Rasters)

Terrain models appear as flat, image-like layers that you can overlay in the viewer. They are useful for understanding the shape of the ground and the height of vegetation.

To view a terrain model:

1. Check its checkbox in the project tree.
2. If multiple terrain types exist (DTM, DSM, CHM, Slope Map, TIN), use the radio buttons to select which one to display — only one can be shown at a time.

Each terrain model comes with a hillshade or colorized version — a visually enhanced image that uses simulated sunlight or color gradients to make the data easier to interpret.

Hover over the info icon next to any terrain model in the tree to see a detailed description of what it shows and how it was created.

For a detailed explanation of each terrain model type, see Terrain Models.

Vector Layers

Vector layers are lines, points, and polygons that represent specific features. They are overlaid on top of the point cloud in the viewer.

To view a vector layer:

1. Check its checkbox in the project tree.
2. The features will appear as colored shapes overlaid on your data.

You can show multiple vector layers simultaneously — for example, contour lines and building footprints together.

The vegetation findings from the power line analysis are grouped under a single Encroachment Findings header, with one severity-colored layer per analysis. Clicking a finding polygon in the viewer opens a panel with its details (cluster ID, points, area, volume, distance to the wire, wire ID).

For a detailed explanation of each vector type, see Vectorization.

Reports

Reports appear at the bottom of the project tree. They are downloadable documents (PDF, CSV, DXF) that package your results into professional deliverables.

Reports cannot be "viewed" in the 3D viewer — they are meant to be downloaded and opened in the appropriate application (PDF viewer, spreadsheet program, or CAD software).

For a detailed explanation of each report type, see Reports.

Comparing Before and After

A useful workflow after processing is to compare your original data with the classified results:

1. Toggle between original and classified point clouds — untick one and tick the other to see the difference.
2. Switch visualization modes — go from RGB (original colors) to Classification (AI-assigned labels) and back.
3. Overlay terrain models on the point cloud — tick a DTM or CHM layer to see the terrain surface alongside the points.
4. Add background satellite imagery — click the Background Map button to compare your data with satellite photos.

What If Results Look Wrong?

The AI classification is typically 95%+ accurate for common classes (ground, buildings, vegetation), but no automated system is perfect. Common issues include:

Issue	What to Do
Some ground points classified as vegetation	Use the Classification Editing tools to manually correct them
Building edges misclassified	Normal for complex roof shapes — can be corrected manually
Wet roads or dark surfaces classified as Water	Enable water extraction — stray water labels are cleaned up automatically when water is extracted
A bridge deck showing as Ground or Road	Enable bridge extraction — the deck is relabeled automatically from the bridge polygon
Some classes are empty	Your data may not contain those features (e.g., no power lines in a forest survey)
Noise points visible	Untick "Low Point" and "High Noise" classes to hide them
Terrain model has holes	Occurs in areas with very sparse ground points — try a coarser resolution

Next Step

For a deeper understanding of each output product, continue to:

• Terrain Models — DTM, DSM, CHM, Slope, TIN in detail
• Vectorization — Contours, buildings, bridges, trees, water, power lines in detail
• Reports — Professional reports and documents

Or, to learn about the interactive tools, skip to Measurement Tools.