LiDAR or Photogrammetry? Choosing the Right Drone Payload for a Site Survey
Both produce survey-grade data, but they fail in different places. How to choose between LiDAR and photogrammetry, and what actually drives accuracy.
ByPremesh Engineering Team· Engineering
Published 2 min read

Key takeaways
- LiDAR can record ground returns beneath gaps in vegetation and works on low-texture surfaces; photogrammetry gives rich colour and is usually cheaper.
- Accuracy depends as much on ground control, GNSS corrections and flight planning as on the sensor.
- Decide from the deliverable backwards: terrain model, volumes, progress imagery or as-built comparison.
- Check local drone regulations before planning flights.
"Which is more accurate?" is the wrong first question. LiDAR and photogrammetry can both deliver survey-grade results when the workflow is right, and both can disappoint when it is not. The better question is what do you need to deliver, and what does the site look like?
How they differ
Photogrammetry reconstructs 3D geometry from overlapping photographs. It produces high-resolution orthomosaics and textured models, which are excellent for progress documentation, visual inspection and volume calculations on open ground. It struggles where surfaces lack texture (smooth concrete, water, uniform gravel), where lighting changes between photos, and where vegetation hides the ground.
LiDAR measures distance with laser pulses and records multiple returns. Some pulses pass through gaps in vegetation and reach the ground, which makes LiDAR the usual choice for bare-earth terrain models under cover. It works in lower light and on low-texture surfaces, but the payload is more expensive and the data needs more processing.
What really drives accuracy
- Ground control. Surveyed control points, or a well-managed RTK or PPK workflow, anchor the model to a known coordinate system. Skipping control is the most common cause of disappointing accuracy.
- Flight design. Altitude, overlap, speed and line spacing set the resolution and the failure tolerance of the dataset.
- Site conditions. Wind, sun angle, standing water and moving plant all affect results.
- Processing and checks. Verify against independent check points that were not used in the adjustment.
A simple way to choose
| If you need | Lean towards |
|---|---|
| Progress imagery, visual records, volumes on open ground | Photogrammetry |
| Terrain beneath vegetation, corridor mapping | LiDAR |
| Both a visual record and a terrain model | Combined RGB + LiDAR payload |
Output formats matter
Agree the deliverables before flying: point cloud format, coordinate system and datum, DEM or DTM resolution, and whether BIM-ready outputs are required. It is much cheaper to specify these up front than to reprocess later.
Regulations
Rules for flying drones commercially differ by country and sometimes by site. Check the requirements of the relevant authority, for example the FAA in the United States or the Ministry of Civil Aviation and DGCA in India, and confirm permissions before mobilising.
References & further reading
- Unmanned Aircraft Systems (UAS) — U.S. Federal Aviation Administration
- Ministry of Civil Aviation, India — Government of India
Editorial note: General guidance only: always follow your project engineer of record and local regulations. Draft prepared with AI assistance and pending technical review by a Premesh engineer.
- #Drone
- #LiDAR
- #Surveying
