TL;DR
An RGB imagery survey captures high-resolution, true-colour photographs of a site — almost always from a drone — using a standard red-green-blue camera sensor. Those overlapping images are processed with photogrammetry to produce measurable deliverables: orthomosaics, point clouds, surface models and contours, all georeferenced to real-world coordinates.
Key takeaways
- An RGB imagery survey records visible-light colour photographs (the same red, green and blue channels the human eye sees) and converts them, through photogrammetry, into scale-correct maps and 3D models.
- A typical drone RGB survey resolves at a ground sampling distance of 1-3 cm/pixel and is georeferenced to GDA2020 / MGA2020 with AHD heights, the Australian national datum.
- With RTK or PPK positioning and surveyed ground control, ISS delivers RGB-derived products accurate to roughly 2-5 cm horizontally — fit for stockpile volumes, as-built overlays and compliance records.
- RGB differs from multispectral (extra near-infrared/red-edge bands for vegetation health) and LiDAR (laser ranging that sees through canopy); RGB is the lowest-cost, highest-resolution way to capture visual detail and 3D geometry of open ground.
- Indicative Australian cost is AUD 2,000-3,500 per drone day plus processing — well below manned aerial photography for sites under roughly 200 ha.
What is an RGB imagery survey?
Definition: An RGB imagery survey is the collection of overlapping true-colour (red-green-blue) photographs across a site, captured by a camera — typically drone-mounted — and processed photogrammetrically into measurable, georeferenced spatial products such as orthomosaics, dense point clouds and digital surface models.
"RGB" simply names the three colour bands a conventional digital camera records. An RGB sensor captures the visible spectrum the same way your phone or a DSLR does, which is why an RGB image looks natural to the eye. The survey distinction is not the camera type but what happens next: the photographs are flown in a planned grid with deliberate overlap so that every point on the ground appears in several frames from slightly different angles.
That overlap is what makes the imagery measurable. Photogrammetry software identifies common features across the images and uses the parallax between viewpoints to reconstruct three-dimensional position — the same principle the human brain uses to perceive depth from two eyes. The output is no longer a stack of pretty pictures but a coordinated dataset you can take real measurements from.
What makes RGB the workhorse of drone surveying is resolution per dollar. A modern full-frame RGB sensor packs far more usable detail into each pixel than a multispectral array, and it carries no laser unit, so the payload is lighter, the flights are longer and the data is cheaper to process. For mapping bare ground, stockpiles, earthworks and built infrastructure, an RGB survey is usually the most cost-effective starting point.
How an RGB imagery survey works
The RGB survey workflow runs from flight planning to georeferenced deliverables, typically over one to three days of fieldwork for a site survey.
Flight planning and control: The survey boundary, flight altitude and image overlap (around 75% front, 65% side) are set to hit the target ground sampling distance. Ground control points are placed and surveyed with a GNSS receiver to tie the imagery to GDA2020 / MGA2020 coordinates and AHD heights.
Image capture: An RTK drone — commonly a DJI Matrice 350 RTK carrying a Zenmuse P1 (45 MP full-frame RGB) — flies the programmed grid, recording hundreds to thousands of overlapping true-colour images, each geotagged with a precise camera position.
Aerial triangulation: Photogrammetry software (Pix4Dmapper, Agisoft Metashape or DJI Terra) matches features across the overlapping frames and solves the exact position and orientation of every photo through bundle adjustment.
Point cloud and surface model: A dense, coloured point cloud is generated, then gridded into a digital surface model (DSM) or filtered to a bare-earth digital terrain model (DTM) for elevation and volume work.
Orthomosaic and derivatives: Each image is orthorectified and blended into a single scale-correct orthomosaic; contours, breaklines and CAD surfaces are extracted as required.
Delivery and QA: Products are exported (GeoTIFF, LAS/E57, DXF) with a quality report checking accuracy against independent survey points not used in processing.
Key point: RGB photogrammetry only reconstructs surfaces the camera can see. Dense vegetation, deep shadow and water defeat it, and the elevation model is only as good as the overlap and control behind it — which is why ISS plans flight lines and ground control to the deliverable, not the other way around.
RGB imagery survey vs multispectral vs LiDAR
RGB is one of three common drone sensing methods. The right choice depends on what you need to measure and what surface you are mapping.
| Aspect | RGB imagery | Multispectral | LiDAR |
|---|---|---|---|
| What it records | Visible true colour (R, G, B) | Visible plus near-infrared / red-edge bands | Laser range returns |
| Resolution | Highest (1-3 cm/px typical) | Lower per band | Point density, not pixels |
| Sees through vegetation | No | No | Partially (canopy penetration) |
| 3D geometry | Yes, via photogrammetry | Yes, but coarser | Yes, directly measured |
| Best for | Mapping, stockpiles, as-builts, visual records | Crop/rehab vegetation health | Vegetated terrain, dense plant, powerlines |
| Relative cost | Lowest | Moderate | Highest |
In practice, RGB is the default for open mine pits, construction earthworks and asset documentation. Multispectral is added when vegetation condition matters — for example monitoring mine rehabilitation. LiDAR earns its higher cost where canopy or dense steelwork hides the ground from a camera.
How accurate is an RGB imagery survey?
Accuracy is driven by ground sampling distance (GSD), positioning quality and the density of ground control. As a rule of thumb, horizontal accuracy is roughly one to two times the GSD.
| Method | GSD | Horizontal accuracy | Notes |
|---|---|---|---|
| RTK/PPK drone + ground control | 1-2 cm/px | 2-3 cm | Best practice for engineering and as-built work |
| RTK drone, minimal control | 2-3 cm/px | 3-5 cm | Standard for most mine and construction sites |
| Drone, no RTK, control only | 3-5 cm/px | 5-10 cm | Budget mapping, lower precision |
| Manned aerial RGB survey | 7-15 cm/px | 15-40 cm | Regional and large-corridor coverage |
Vertical accuracy from RGB photogrammetry is typically a little looser than horizontal — commonly 1.5 to 3 times the GSD on well-controlled flights. Where a project demands tighter vertical tolerances, ISS strengthens the control network and validates against independent check points before sign-off.
Where RGB imagery surveys are used in industry
ISS captures RGB imagery surveys across mining, civil construction and resources operations Australia-wide, from the Pilbara iron ore hubs around Port Hedland and Newman to the Bowen Basin coalfields near Moranbah and Blackwater.
Mining and resources
RGB orthomosaics underpin monthly open-pit mapping, stockpile volume verification, haul-road condition reviews and progressive rehabilitation records. A dated, georeferenced RGB capture of a tailings storage facility or waste dump provides defensible evidence for compliance reporting to state regulators.
Civil construction and earthworks
RGB-derived surfaces drive cut-and-fill volumes and as-built comparison against design, while the orthomosaic documents construction progress for payment claims and dispute records. A single flight can map a 50-hectare site in under an hour and return a measurable colour image the same day.
Asset management and environment
Georeferenced RGB imagery feeds GIS and CAD as a real-world base layer for feature location, change detection between captures, erosion monitoring and asset inspection across ports, solar farms and linear corridors. The natural-colour record is also the most readable deliverable for non-technical stakeholders.
All ISS UAV capture is conducted under CASA Part 101 rules, flown under the appropriate remotely piloted aircraft operator's certificate with licensed pilots.
Equipment and software
| Component | Typical ISS setup | Purpose |
|---|---|---|
| Drone platform | DJI Matrice 350 RTK / Mavic 3 Enterprise | Stable capture with RTK positioning |
| RGB camera | Zenmuse P1 (45 MP full-frame) | High-resolution true-colour imagery for fine GSD |
| Positioning | RTK/PPK + CORSnet or local base | Centimetre-level camera positions |
| Ground control | Surveyed checkerboard targets (GNSS to MGA2020 / AHD) | Georeferencing and accuracy verification |
| Processing | Pix4Dmapper, Agisoft Metashape, DJI Terra | Triangulation, point cloud, surface model, orthomosaic |
Survey-grade ground control is established with GNSS receivers and, where required, Leica or Trimble total stations to validate deliverables against independent check points. For dense plant and indoor areas where flight is impractical, terrestrial FARO or Leica laser scanners supply geometry the camera cannot reach.
Frequently asked questions
What is an RGB imagery survey?
An RGB imagery survey is the capture of overlapping true-colour (red-green-blue) photographs of a site, usually by drone, which are then processed with photogrammetry into measurable, georeferenced products such as orthomosaics, point clouds and surface models. It is the most common and cost-effective form of drone mapping for open ground and built infrastructure.
What does RGB mean in surveying?
RGB stands for red, green and blue — the three visible colour channels a standard digital camera records. An RGB survey uses a conventional colour sensor, as distinct from a multispectral sensor (which adds invisible near-infrared bands) or a LiDAR unit (which uses a laser rather than a camera).
How accurate is an RGB imagery survey?
A drone RGB survey flown with RTK positioning and ground control typically achieves 2-5 cm horizontal accuracy at a ground sampling distance of 1-3 cm/pixel, with vertical accuracy a little looser. Accuracy depends on flying height, image overlap, positioning quality and the number and spread of ground control points.
What is the difference between RGB and multispectral or LiDAR surveys?
RGB records visible colour at the highest resolution and lowest cost, ideal for mapping, stockpiles and as-builts. Multispectral adds near-infrared bands to assess vegetation health. LiDAR uses a laser that can partly penetrate vegetation, making it the choice for densely vegetated terrain or cluttered plant where a camera cannot see the ground.
How much does an RGB imagery survey cost in Australia?
Drone RGB capture is typically AUD 2,000-3,500 per field day, with total project cost set by site size, accuracy and deliverables. A small site (up to 10 ha) commonly runs AUD 3,000-6,000 including processing, while large sites (50-200 ha) range AUD 8,000-20,000 — usually well below traditional manned aerial photography for the same coverage.
Request a quote
If you need measurable, true-colour mapping of your site — for stockpile volumes, as-built overlays, compliance records or progress monitoring — an RGB imagery survey is the fastest, most cost-effective place to start. ISS flies CASA Part 101 RTK drone surveys, ties every deliverable to GDA2020 / MGA2020 and AHD, and returns georeferenced orthomosaics, point clouds and surface models with a verified accuracy report. Call ISS on 0407 057 015 to discuss your RGB imagery survey requirements and receive a fixed-price quote.