TL;DR
UAV LiDAR fires laser pulses to measure ground directly, so it sees bare earth through vegetation and works in poor light — but the payloads are dear and the deliverable has no native colour. UAV photogrammetry reconstructs 3D geometry from overlapping photos, giving photorealistic orthomosaics at roughly half the field cost, provided the surface is exposed and well textured. On open Pilbara stockpiles either works; over rehabilitated waste dumps or scrubby haul-road corridors, LiDAR is usually the only honest answer.
Key takeaways
- On bare, exposed surfaces both methods hit comparable vertical accuracy — around 2–5 cm with proper ground control; the gap only opens once vegetation enters the frame, where UAV LiDAR holds 2–5 cm to ground while photogrammetry drifts to 10–50 cm because it can only see the canopy top.
- A typical 50 ha UAV photogrammetry capture runs roughly AUD 2,500–6,000; the equivalent UAV LiDAR job is AUD 6,000–12,000, driven almost entirely by the AUD 80k–250k sensor cost (a DJI Zenmuse L2 sits near the low end, a RIEGL miniVUX or YellowScan payload near the top).
- Both are commercial RPA operations: under CASA Part 101 you need a Remote Pilot Licence (RePL) and the operating entity an RPA Operator's Certificate (ReOC), since survey-grade platforms almost always exceed the 25 kg / sub-2 kg excluded thresholds.
- Photogrammetry struggles with water, fresh white surfaces and uniform sand because it needs texture to match pixels; LiDAR struggles with water, very dark bitumen and active dust because the pulse is absorbed or scattered.
- Neither replaces terrestrial work: at 2–5 cm absolute, UAV data cannot certify a turbine baseplate or mill trunnion to ±2 mm — that stays with the total station and tripod scanner.
How UAV LiDAR works
UAV (or RPAS) LiDAR is active remote sensing. The airborne sensor emits 100,000–2,000,000 laser pulses per second and times each return; combined with a GNSS receiver and an inertial measurement unit logging attitude at 200–1,000 Hz, the system resolves a georeferenced point cloud in GDA2020 / MGA2020 with AHD heights.
The decisive trait is multiple returns. A single pulse fired into mulga or buffel grass can return first off the leaf, again off a branch, and finally off the soil. Classifying those returns lets the surveyor build both a digital surface model (top of cover) and a true bare-earth digital terrain model — the deliverable that hydrologists and earthworks engineers actually need.
| UAV LiDAR parameter | Entry payload | Professional payload | Survey-grade payload |
|---|---|---|---|
| Absolute vertical accuracy | 5–15 cm | 2–5 cm | 1–3 cm |
| Absolute horizontal accuracy | 10–30 cm | 5–10 cm | 3–5 cm |
| Point density | 50–100 pts/m² | 100–300 pts/m² | 300–1,000+ pts/m² |
| Returns per pulse | 1–3 | 3–5 | 5–15 |
| Representative sensor | DJI Zenmuse L2 | YellowScan Mapper+ | RIEGL miniVUX-3 |
How UAV photogrammetry works
UAV photogrammetry is passive. The aircraft flies a programmed grid — typically 70–80% forward and 60–70% side overlap — capturing hundreds to thousands of RGB frames. Software then performs aerial triangulation to recover each camera station, dense-matches pixels across overlapping frames, and outputs a coloured point cloud, a textured mesh, and a georeferenced orthomosaic.
The output is its strength: the model carries real photographic texture, so it reads instantly for non-technical stakeholders — a mine manager can see the actual berm, not an abstract point cloud. Accuracy hinges on ground control. With well-distributed surveyed ground control points (GCPs), or an RTK/PPK aircraft such as a DJI Matrice 350 RTK with a Zenmuse P1, photogrammetry reaches centimetre absolute accuracy. Without control it can drift by metres.
| UAV photogrammetry parameter | No GCPs | With GCPs | RTK/PPK aircraft |
|---|---|---|---|
| Horizontal accuracy | 1–5 m | 1–3 cm | 1–3 cm |
| Vertical accuracy | 2–10 m | 2–5 cm | 2–4 cm |
| Ground sample distance | 1–5 cm/pixel | 1–5 cm/pixel | 1–5 cm/pixel |
| Native colour | Yes | Yes | Yes |
| Canopy penetration | None | None | None |
Ground control points (GCPs) are physically surveyed targets visible in the imagery. They anchor the photogrammetric adjustment to a known datum. RTK/PPK positioning on the aircraft reduces — but rarely fully eliminates — the need for GCPs on accuracy-critical jobs.
Accuracy and capability: side by side
| Metric | UAV LiDAR | UAV photogrammetry | Practical read |
|---|---|---|---|
| Bare-earth vertical | 2–5 cm | 2–5 cm (with control) | Even on exposed ground |
| Vegetated terrain vertical | 2–5 cm to ground | 10–50 cm (canopy only) | LiDAR's decisive edge |
| Hard-surface detail | 1–3 cm | 1–3 cm | Both excellent |
| Edge / breakline definition | Excellent | Good (softens at discontinuities) | LiDAR sharper on batters |
| Low light / shadow | Works (active) | Fails (needs even light) | LiDAR for dawn/dusk windows |
| Native deliverable | Point cloud, DTM/DSM | Orthomosaic, mesh, coloured cloud | Photogrammetry for visuals |
| Dust tolerance | Better | Poorer | Matters near active crushing |
Key point: above roughly 30% canopy cover where you still need ground levels, UAV photogrammetry cannot deliver a defensible bare-earth surface — the pixels simply never see the dirt. That single fact decides most rehabilitation and floodplain jobs.
Cost comparison: 2026 Australian rates
| Cost component | UAV LiDAR | UAV photogrammetry |
|---|---|---|
| Payload + aircraft capital | AUD 80,000–250,000 | AUD 15,000–60,000 |
| Contractor field rate | AUD 3,500–6,000 / day | AUD 1,500–3,500 / day |
| Processing software (annual) | AUD 5,000–20,000 | AUD 500–5,000 |
| GCP establishment | Often not required (RTK/PPK) | AUD 500–2,000 |
| Typical 50 ha project | AUD 6,000–12,000 | AUD 2,500–6,000 |
The LiDAR premium is real and it is almost all hardware: the sensor is five to ten times dearer, insurance and maintenance follow, and the operator training is more specialised. It is justified when the site is vegetated, when multiple returns add value (canopy height, powerline clearance over a mine corridor), or when a tight weather window rewards LiDAR's light-independent, single-pass capture.
Use-case scenarios
Open-pit stockpile volumetrics, Pilbara iron ore
Recommendation: either, slight lean to photogrammetry. Exposed ore and processed product give superb texture; the coloured orthomosaic is easy for a port-side ore controller to read. LiDAR pulls ahead only when conveyor dust is suppressing pixel matching near the crusher.
Rehabilitation and waste-dump monitoring, Bowen Basin coal
Recommendation: UAV LiDAR. Regrowth on rehabilitated dumps demands both bare-earth terrain (for erosion and settlement) and canopy structure (for revegetation compliance). Only multi-return LiDAR captures both in one flight — the standard for environmental reporting against mine-closure conditions.
Haul-road and corridor design through scrub
Recommendation: UAV LiDAR. Mulga and spinifex along an alignment defeat photogrammetry's ground detection. A clean bare-earth DTM is the entire point of a corridor design survey.
Construction progress and earthworks volumes, cleared civil site
Recommendation: UAV photogrammetry. Bare formation and exposed batters reconstruct cleanly, and the orthomosaic doubles as a dated progress record. The 40–60% cost saving compounds across fortnightly captures.
Processing-plant as-built, complex steelwork
Recommendation: neither — use terrestrial laser scanning. Pipes, conveyors and structure need sub-centimetre accuracy for clash detection; UAV data at 2–5 cm cannot support retrofit design. Drone photogrammetry can add colour, but the geometry must come from the ground.
The decision table
| If your priority is… | Choose | Because |
|---|---|---|
| Bare earth under >30% vegetation | UAV LiDAR | Canopy penetration via multiple returns |
| Lowest cost on an exposed site | UAV photogrammetry | ~50% lower field and capital cost |
| Photorealistic visuals for stakeholders | UAV photogrammetry | Native RGB texture |
| Dawn/dusk or shaded capture windows | UAV LiDAR | Light-independent active sensor |
| Sharp batter and breakline definition | UAV LiDAR | Cleaner edges, no matching blur |
| Both terrain and a colour orthomosaic | Both, fused | LiDAR geometry + photogrammetric texture |
| ±2 mm dimensional control | Neither | Terrestrial total station or scanner |
Combining both on one platform
Dual-payload aircraft increasingly carry a LiDAR sensor and an RGB camera together, and the smart workflow uses each for what it does best. Capture both on the same flight; process the LiDAR into the bare-earth DTM and primary cloud; process the imagery into the orthomosaic; register the two on shared control; then colourise the LiDAR cloud and drape the orthomosaic onto LiDAR-accurate terrain. The result carries LiDAR's penetration and edge fidelity with photogrammetry's readability. It costs more than either alone but removes the compromise — common now on large WA and QLD mine captures where one mobilisation has to serve survey, environmental and operations teams at once.
Honest limitations
UAV LiDAR: water and very dark bitumen absorb the pulse and leave gaps; point clouds are large (50 GB+ on a sizeable job) and need real IT to handle; raw data has no colour; capital and operating costs stay high.
UAV photogrammetry: no canopy penetration, full stop; it needs consistent diffuse light, so harsh shadow and glare create artefacts; uniform low-texture surfaces (fresh sand, white walls, calm water) reconstruct poorly; and it is more sensitive to wind, dust and rain at capture time.
Both are also bound by the same airspace reality: many remote operations sit in uncontrolled airspace and proceed simply, but sites near aerodromes such as Port Hedland, Karratha or a busy mine airstrip require coordination with Airservices Australia, and gusts above roughly 35 km/h ground most survey aircraft regardless of payload.
Frequently asked questions
Is UAV LiDAR more accurate than UAV photogrammetry?
On open, well-textured ground, no — both reach about 2–5 cm vertical with proper control. The advantage is situational: over vegetation LiDAR holds that accuracy to the true ground while photogrammetry measures only the canopy top, so error there can be 10–50 cm.
Do I still need ground control points with UAV LiDAR?
If the platform has RTK or PPK GNSS, control points are usually not needed for topographic accuracy, but independently surveyed check points are still strongly recommended on every job for quality verification — and essential where the data must satisfy a compliance or design specification.
Which is better for stockpile volumes?
Either suits exposed stockpiles. Photogrammetry gives a clearer visual for operations staff; LiDAR is marginally better in dusty, active areas and captures slightly faster. The volume figures themselves are comparable when both are properly controlled.
Can UAV LiDAR or photogrammetry replace a total station survey?
No. At 2–5 cm absolute accuracy, neither can certify dimensional control, machine alignment or setting-out work that requires ±2–5 mm. Those tasks stay with total stations and terrestrial scanners; the UAV handles the broad-area topography around them.
What software processes the data?
Photogrammetry typically runs through Pix4D, Agisoft Metashape, DJI Terra or Bentley iTwin Capture. LiDAR processes in vendor tools (DJI Terra for the Zenmuse L2, YellowScan CloudStation) plus LAStools, TerraSolid or CloudCompare for classification and bare-earth extraction.
Ready to choose the right method?
The wrong sensor over the wrong site wastes a mobilisation and produces data you cannot use — a photogrammetry flight over regrowth that never sees the ground, or a costly LiDAR job on a bare pad where photos would have done. Industrial Spatial Solutions runs both UAV LiDAR and photogrammetry payloads across mining and civil sites Australia-wide, and we recommend the method your site and specification actually need, not the kit we happen to prefer. For projects demanding sub-centimetre accuracy we also offer terrestrial 3D laser scanning. Call us on 0407 057 015 to talk through your site, accuracy requirement and budget, and we will quote the right approach — or the right combination.