+91 98765 43210 | info@asav.com | Mon – Sat, 09:00 – 19:00 IST
Noida, Uttar Pradesh 201301
  1. Home
  2. Services
  3. LiDAR
  4. Bare Earth Classification
LiDAR

Bare Earth Classification

Ground extraction, hydro-flattening and DTM production verified against control.

Bare Earth Classification — representative deliverable
Overview

Bare Earth Classification

Bare-earth classification is the most requested LiDAR deliverable and the easiest to get subtly wrong. Automated ground filters do well on open terrain and struggle exactly where accuracy matters most — dense low vegetation, steep slopes, walls and embankments, bridge decks, and the transition between built and natural surfaces.

We run automated classification and then review it systematically in profile, terrain type by terrain type. Ground points wrongly removed from beneath vegetation are recovered; low vegetation, vehicles and temporary structures wrongly retained as ground are removed. Hydro-flattening is applied with water surfaces set at or below their inflow, and breaklines are captured where the terrain demands them.

Vertical accuracy is then verified against control points that were not used in the process, reported per land-cover type because a single aggregate RMSE conceals the categories where the data is weakest.

What you receive

Every item below is issued in your own template, with your layer conventions, title block and revision scheme.

  • Classified point cloud — ground, low / medium / high vegetation, building, water, bridge, noise
  • Bare-earth ground surface as TIN or grid at the specified resolution
  • Hydro-flattened water surfaces with enforced downstream monotonicity
  • Breaklines — hydrographic, road, top and bottom of bank
  • Contours derived from the conditioned bare-earth surface
  • DTM raster at the specified cell size
  • Vertical accuracy report against independent control, by land-cover class
  • Density, void and overlap consistency analysis
  • Tile index and delivery manifest with per-tile QC status

Production detail

Key production parameters for Bare Earth Classification
Typical turnaround1,000–2,000 km² per month per team
Classification standardASPRS LAS class codes, or your project schema
Vertical accuracyVerified to project NVA / VVA; typically 5–15 cm RMSEz
Point density0.5 to 50 pts/m² handled
Hydro-flatteningPer USGS LiDAR Base Specification or your equivalent
Output formatsLAS 1.2 / 1.4, LAZ, COPC, GeoTIFF, DWG, LandXML, SHP breaklines
RevisionsReprocessed until the specification is met

Software and formats

We work in your toolchain. If something you use is not listed, ask — the list below is only what we use most.

TerraScan / TerraModelerLP360Global MapperArcGIS ProCloud CompareLAStoolsPython / PDAL

Frequently asked

Carefully, and with manual review, because it is the hardest case. Automated filters either strip genuine ground or retain vegetation, and which one they do depends on the parameters. We review these areas in profile and adjust locally rather than tuning a single global parameter set across the whole block.

It depends on the acquisition, not on the processing. With good airborne data we typically verify 5–15 cm RMSEz in open terrain, with vegetated categories reported separately because they are always worse. We report against your control and will tell you plainly if the acquisition cannot support the spec.

Yes — hydrographic breaklines for hydro-flattening, plus road, kerb, and top and bottom of bank where the specification or the terrain requires them. Breaklines are what make the difference between a usable surface and a smoothed approximation.

Yes. We work with data from all the common airborne, mobile and terrestrial platforms, in LAS, LAZ or vendor formats, and we handle trajectory data where strip adjustment is part of the scope.

Per-tile classification review status, vertical accuracy against independent control by land-cover class, density and void analysis, overlap consistency between flight lines, and a list of every issue found and how it was resolved.
Step 1 of 4 in LiDAR

Send one bare earth classification job and see how we work

Scope, price and programme come back in writing before anything starts. Most first projects are a single deliverable so you can judge us on output, not on a pitch.

Call an engineer