Point Cloud & 3D Mesh Capture

Existing site conditions, captured as 3D point cloud and textured mesh.

Orbital captures the existing site — buildings, hard surfaces, roads, vegetation, terrain — as both a high-resolution 3D point cloud and, where the project needs it, a fully textured 3D mesh model. Deliverables land in your engineering, BIM, GIS or reality-modelling workflow in the format your platform expects — Bentley iTwin / ContextCapture, Autodesk Revit / Civil 3D / ReCap, Esri ArcGIS Pro, SuperMap, or any of the open-standard tools that accept the formats below.

Two distinct deliverable types — point cloud and mesh

Reality capture produces two related-but-different outputs. Most engineering projects use one or the other; some use both. Knowing which you need is the first scoping question.

Point cloud

Raw 3D spatial data — tens of millions of individual XYZ points across the captured area, each tagged with RGB colour from the photogrammetry capture. Best for: dimensional verification, design-model reference, surface measurement, volumetric calculation, integration into BIM as a referenced cloud, GIS analysis. The point cloud is the source dataset; downstream tools can derive surfaces, CAD geometry or measured features from it.

3D mesh model

A continuous textured 3D surface model derived from the point cloud and photographic capture. Looks like a navigable 3D representation of the real site. Best for: visualisation, client presentation, design context, planning and approval submissions, integration into reality-modelling platforms like Bentley iTwin, Esri ArcGIS Pro Scene Layers, or BIM/AEC visualisation tools. The mesh is the rendered product; it’s what non-technical stakeholders can read intuitively.

Point cloud — deliverable formats

Orbital point cloud deliverables are available in the following formats, chosen to match your downstream platform:

  • .las — the ASPRS open standard. Universal across BIM, GIS, survey and engineering tools. Default deliverable format unless specified otherwise.
  • .laz — compressed LAS format. Same data, around 80% smaller files. Useful for large captures or when distributing across multiple project stakeholders.
  • .ply — Stanford Polygon format. Used in computer graphics, research and several visualisation pipelines. Also handles textured meshes.
  • .pcd — Point Cloud Data, native to the open-source Point Cloud Library (PCL). Used in robotics, computer vision research and AI / machine-learning workflows.
  • .s3mb — SuperMap 3D Model Block. For organisations standardised on the SuperMap iServer / iEarth platform.

Every point cloud is georeferenced to GDA2020 Zone 56 using surveyed ground control captured with our GNSS RTK kit at the time of flight, so the dataset sits in absolute coordinates rather than only relative to itself.

3D mesh — deliverable formats

Where the project calls for a textured mesh model rather than (or in addition to) the raw point cloud, we deliver in the following formats:

  • .osgb — OpenSceneGraph Binary. Native format for Bentley ContextCapture and iTwin reality-modelling platforms. The standard output for large-scale infrastructure and city-scale reality modelling.
  • .obj — Wavefront. Universal mesh format supported by virtually every 3D tool — from CAD packages through to visualisation, animation and game engines. The safe default when downstream compatibility is uncertain.
  • .fbx — Autodesk’s animation / 3D asset format. Imports directly into Revit, 3ds Max, Maya, Unity, Unreal Engine. The format of choice for BIM-led projects with a visualisation component.
  • .i3s — Esri’s Indexed 3D Scene format (now an OGC open standard). For projects standardised on ArcGIS Pro Scene Layers and SLPK packaging.
  • .s3md — SuperMap 3D Model Data. Mesh format for the SuperMap platform.
  • .ply — Stanford Polygon (handles both points and meshes). Useful where the downstream pipeline accepts PLY natively.

Mesh deliverables include the texture maps alongside the geometry — so the model appears in its real colours and surface detail when loaded, not as a grey untextured shape.

Which format do I need?

If you’re unsure of the format your downstream platform expects, this is the quick reference:

Bentley workflows (iTwin, ContextCapture, MicroStation)

Mesh: .osgb (native), .obj or .fbx (universal fallback). Point cloud: .las or .laz.

Autodesk workflows (Revit, Civil 3D, AutoCAD, 3ds Max)

Mesh: .fbx (best Revit integration), .obj (universal). Point cloud: .las or .laz.

Esri / GIS workflows (ArcGIS Pro)

Mesh: .i3s (SLPK). Point cloud: .las or .laz.

SuperMap workflows

Mesh: .s3md. Point cloud: .s3mb.

Visualisation, animation or game-engine workflows

Mesh: .fbx or .obj. Point cloud: .ply or .las.

Research, robotics or machine learning

Point cloud: .pcd or .ply. Mesh: .ply or .obj.

If your platform isn’t covered above, tell us at scoping. Most professional engineering and visualisation tools accept at least one of the formats we deliver natively, and where translation is required between formats it’s a quick post-processing step.

How we capture

Orbital’s point cloud deliverables are built around visible-light drone capture and supported by GNSS RTK and total station geometry capture. The drone produces the dense surface reconstruction; the GNSS and total station work tie the result into a survey-grade coordinate frame and capture the discrete features that matter at higher accuracy than the drone produces by itself.

Visible-light drone photogrammetry

A UAV flies a planned flight grid above the site. Hundreds of overlapping high-resolution photographs feed a photogrammetry pipeline that reconstructs the 3D geometry and texture of every visible surface. Result: dense, colour-accurate point cloud and mesh across the captured area.

Photogrammetry is well-suited to:

  • Outdoor sites — buildings, hard surfaces, infrastructure, low vegetation, terrain.
  • Colour-accurate capture — RGB texture from the source photography flows through to both point cloud and mesh deliverables.
  • Combined deliverables from a single flight — point cloud, mesh, drone orthomosaic and DEM all produced from the same source data.

GNSS RTK ground control

Surveyed ground control points are placed across the site before flight and captured to GDA2020 Zone 56 with our GNSS RTK kit. These provide the survey-grade reference frame that distinguishes engineering-grade reality capture from a casual drone overflight, and they’re what gives the resulting cloud and mesh their absolute accuracy.

Total station geometry capture

Where specific features need to be captured at accuracy beyond the drone’s typical performance — small drainage features, services exposed at ground level, critical infrastructure points, edges that matter to the design — we capture them with total station as part of the same engagement. The combined dataset gives drone-grade coverage of the broad site with sub-centimetre accuracy on the elements that need it.

Where reality capture earns its place

Adaptive reuse and retrofit design

Existing buildings rarely match their original drawings. Point cloud and mesh capture is the only realistic way to bring actual current geometry into the design model. Particularly relevant for Brisbane’s older commercial stock, heritage buildings and adaptive-reuse projects.

Reality modelling for infrastructure projects

Bridge inspections, tunnel surveys, road corridor capture, industrial facility documentation. The textured mesh from a reality-modelling pipeline (delivered as .osgb for Bentley workflows, .fbx for Autodesk, .i3s for Esri) is the engineering-grade record of the asset’s current state.

BIM existing-conditions models

Architects and BIM coordinators benefit from a point cloud or mesh of the existing site as the basis for the design model. Eliminates the guesswork that legacy as-built drawings introduce.

Civil works planning

Pre-construction terrain, vegetation, hard surface and infrastructure capture. Useful for earthworks design, drainage modelling, design verification and post-construction comparison.

Asset documentation and 3D as-built records

Facility managers, councils and large asset owners increasingly require 3D as-built records as part of project handover. Mesh deliverables in .osgb, .i3s or .fbx provide the long-term spatial record.

Forensic and pre-event documentation

Pre-construction baseline of neighbouring properties, post-event damage documentation, boundary or condition disputes. Date-stamped 3D record.

Integration with the wider Orbital service stack

Reality capture sits naturally alongside our locating, survey and drone work — same coordinate system, same project lead, one coordinated deliverable. A typical integrated scope:

  • Underground service locating (EMF + GPR) to AS 5488 quality levels — what’s beneath the site.
  • Drone photogrammetry orthomosaic — plan-view aerial of the site at capture date.
  • 3D point cloud and textured mesh — full 3D dataset of the above-ground existing conditions.
  • Topographic and contour survey — terrain captured to engineering specification.
  • CAD and BIM deliverable — DWG, PDF, IFC, plus the format-specific reality-modelling outputs (.osgb / .fbx / .i3s) integrating all of the above.

This integrated above-and-below-ground spatial deliverable is the work that defines Orbital as a spatial-intelligence specialist. Few SEQ operators deliver the full stack; we do, and we deliver it in the format your project’s platform actually accepts.

Why Orbital for reality capture

Twenty-three years of survey-grade spatial work in South East Queensland — locating, surveying, capturing and delivering spatial data into engineering workflows since well before reality modelling existed as a named discipline. CERTLOC and NULCA-credentialled foundation. The same survey discipline that produced our locating deliverables now produces our point cloud and mesh deliverables.

We deliver into the platforms engineering buyers actually use — Bentley iTwin / ContextCapture, Autodesk Revit / Civil 3D / ReCap, Esri ArcGIS Pro, SuperMap iServer — in the formats those platforms expect. We don’t ask the client to translate. The deliverable is the product, and the product has to land ready for the downstream work.

Frequently asked questions

Do I need point cloud, mesh, or both?

Depends on the work. Point cloud is the better input for dimensional verification, design reference and surface measurement. Mesh is the better deliverable for visualisation, presentation, reality-modelling platforms and any context where non-technical stakeholders need to read the 3D model intuitively. Some projects need both — the cloud for the engineering and BIM workflows, the mesh for the visualisation and presentation tracks. We’ll recommend the right deliverable mix at scoping.

Can you deliver in the format my Bentley / Esri / SuperMap / Autodesk platform needs?

Yes. Point cloud: .las, .laz, .ply, .pcd, .s3mb. Mesh: .osgb (Bentley), .fbx (Autodesk), .i3s (Esri), .s3md (SuperMap), .obj and .ply (universal). If your platform expects something not listed, tell us at scoping — most professional engineering platforms accept at least one of these formats natively, and translation to other formats is generally a quick post-processing step.

When is drone photogrammetry the right method, and when isn’t it?

For outdoor sites with good light and unobstructed surfaces — buildings, infrastructure, hard surfaces, terrain with low vegetation — drone photogrammetry produces dense, colour-accurate point clouds and meshes at engineering grade. Where vegetation is dense enough that we’d need to capture the ground beneath canopy, photogrammetry will capture canopy top rather than ground surface; in those cases we supplement with total station survey of the ground beneath canopy, or scope a ground survey for that zone. We tell you up front at scoping which method fits the site.

How accurate is the deliverable?

Drone photogrammetry with surveyed ground control achieves centimetre-scale absolute accuracy across the captured area. Specific accuracy varies with flight height, ground control density and surface conditions — we quote site-specific accuracy figures with each scope. Where higher accuracy is required on specific features, we combine the drone capture with terrestrial total station survey of those features.

Can the capture be combined with underground services and surveyed terrain?

Yes — that’s exactly what we do. The cloud and mesh are georeferenced to GDA2020 Zone 56, the same coordinate system used for our locating and survey work. We deliver integrated datasets combining above-ground reality capture, below-ground located services (designated to AS 5488 quality levels), terrain contours and orthomosaic — one engagement, one project lead, one CAD / BIM package.

How long does a capture take?

Depends on site size and complexity. A typical commercial site of around 1–2 hectares is one to two flight hours on site, plus a day or two for processing and ground control. Larger sites or higher-detail captures take longer; very small sites can be captured in a single morning. We quote the program with each scope.

What does it cost?

Per scope. Variables: site size, deliverable type (point cloud, mesh, both), level of detail, ground control density, deliverable formats, integration with other Orbital services. Send us the project brief and a site address; we’ll come back with a fixed-price quote, usually within 48 hours given the planning involved.

Related services

  • Drone mapping and orthomosaic
  • Topographic and contour survey
  • Underground service locating
  • AS 5488 Subsurface Utility Engineering
  • CAD drawings — DWG and PDF

Tell us about the site and the platform.

Send us the site address, the project type, and the platform / format your downstream workflow expects. We’ll come back with a fixed-price scope including capture method, deliverable formats, and integration with any locating, survey or AS 5488 work the project also needs.