LiDAR vs Photogrammetry: Which Should You Use?

Use LiDAR when you need to measure terrain beneath vegetation, require consistent point spacing across the site, or are surveying large areas where processing time matters. Use photogrammetry when you need full-colour visual outputs like orthomosaics and textured 3D models, when the site is mostly open, or when budget is the primary constraint. Many survey companies — mine included — use both on the same project, because they complement each other: LiDAR for accurate terrain, photogrammetry for visual context.

This is the most common question I get from clients and from surveyors evaluating their equipment investment. The answer depends on your site conditions, deliverables, budget, and timeline. Here’s the full breakdown.

How each technology works

LiDAR

LiDAR (Light Detection and Ranging) fires rapid laser pulses at surfaces and measures the time each pulse takes to return. A UAV-mounted LiDAR sensor emits 240,000 to 1,500,000 pulses per second, each one generating a precise 3D coordinate. The result is a point cloud — millions of XYZ measurements with intensity values.

Key characteristics:

• Active sensor — generates its own light, so it works in low-light conditions and isn’t affected by shadows.
• Multiple returns — a single pulse can record reflections from canopy, understory, and ground.
• Direct georeferencing — each point is positioned using onboard GPS/INS, with no need for feature matching.
• No colour — standard LiDAR records intensity (reflectance strength), not RGB colour. Some systems have an integrated camera, but the LiDAR data itself is monochrome.

Photogrammetry

Photogrammetry uses overlapping photographs to reconstruct 3D geometry. A drone flies a grid pattern, capturing images with 70-85% overlap. Software identifies matching features across images, calculates camera positions (Structure from Motion), and generates a dense point cloud, mesh, and orthomosaic.

Key characteristics:

• Passive sensor — relies on ambient light, so it’s affected by shadows, overcast conditions, and time of day.
• Full colour — every point in the point cloud inherits RGB values from the source photographs.
• Requires texture — the software needs visual features to match between images. Uniform surfaces (snow, sand, water, fresh asphalt) cause failures.
• Multiple deliverable types — point clouds, textured meshes, orthomosaics, DSMs, and 3D models from the same dataset.

Detailed comparison table

When to use LiDAR

Dense vegetation

This is LiDAR’s defining advantage. Laser pulses pass through gaps in the canopy and record returns from the ground below. Photogrammetry cameras can only see the top of the canopy — they have no way to measure the terrain underneath.

If your site has trees, scrub, or tall grass and you need a bare earth DTM, LiDAR is the only reliable option. I’ve seen photogrammetry-derived DTMs under eucalyptus canopy that were 2-3 metres off the actual ground level. That’s not an error you can calibrate away — the data simply isn’t there.

Large area surveys

LiDAR processing is significantly faster than photogrammetry for large datasets. A 200-hectare LiDAR survey might process in 2-4 hours. The equivalent photogrammetry dataset (thousands of overlapping images) could take 12-48 hours in Metashape or Pix4D, depending on hardware.

For corridor surveys (roads, railways, pipelines, power lines), LiDAR’s narrow swath and fast processing make it the standard choice.

Challenging lighting conditions

LiDAR doesn’t care about shadows, overcast skies, or time of day. Photogrammetry image quality degrades with harsh shadows, flat overcast light (which reduces texture contrast), and low sun angles. If you can only fly at noon in harsh sun or on an overcast day, LiDAR gives you consistent results regardless.

High accuracy requirements

While both technologies can achieve centimetre-level accuracy, LiDAR is more consistently accurate because:

• Direct georeferencing removes the dependency on ground control points.
• Point spacing is uniform regardless of surface texture.
• Accuracy is not affected by surface colour, shadow, or reflectivity (within reason).

For engineering surveys where 2-3 cm vertical accuracy is specified, LiDAR provides higher confidence.

Terrain-focused deliverables

If the primary outputs are a DTM, contour map, cross-sections, and volume calculations, LiDAR is the more direct path. The classified point cloud gives you the ground surface without the ambiguity of photogrammetric filtering.

When to use photogrammetry

Visual deliverables

Photogrammetry produces outputs that LiDAR cannot:

• Orthomosaics — Georeferenced, top-down imagery of the entire site. Essential for progress monitoring, planning overlays, and client communication. See What is an Orthomosaic?
• Textured 3D meshes — Photorealistic 3D models viewable in a browser. See How to Share a 3D Model Online.
• 3D Tiles — Streamable 3D models for web viewing via CesiumJS.

If your client needs to see what the site looks like — not just its geometry — photogrammetry is the answer.

Lower budget

A quality photogrammetry camera (DJI Phantom 4 RTK, DJI Mavic 3 Enterprise) costs $2,000-$8,000. A quality LiDAR sensor (DJI Zenmuse L2, YellowScan Mapper+) costs $15,000-$80,000+. For a small survey company, that cost difference determines viability.

Data collection costs per hectare are also lower for photogrammetry because the equipment and mobilisation costs are lower.

Open, textured sites

On sites with minimal vegetation — construction sites, quarries, urban areas, agricultural land after harvest — photogrammetry performs well. There’s no canopy to penetrate, the surfaces are textured enough for feature matching, and the visual outputs add value.

3D modelling

Photogrammetry generates dense, coloured meshes that represent real-world appearance. LiDAR generates point clouds that represent geometry accurately but lack visual texture. For applications like heritage documentation, marketing visualisation, or virtual site tours, photogrammetry’s visual richness is a major advantage.

Smaller sites

For sites under 10-20 hectares, photogrammetry’s lower equipment cost and simpler workflow make it the practical choice unless vegetation penetration is required. A single drone flight with a good camera, processed in Metashape, delivers a complete dataset.

When to use both

Many projects benefit from combining LiDAR and photogrammetry. This is increasingly common as sensors become more affordable and workflows more streamlined.

Typical hybrid workflow

1. Fly LiDAR for terrain measurement, especially in vegetated areas.
2. Fly photogrammetry (often on the same flight with a dual-sensor drone, or as a separate mission) for orthomosaic and visual outputs.
3. Merge the photogrammetry colour data onto the LiDAR point cloud for a colourised point cloud with accurate geometry.

When hybrid makes sense

• Construction monitoring — LiDAR DTM for volume calculations, orthomosaic for visual progress reporting.
• Forestry — LiDAR for canopy height model and terrain, photogrammetry for species identification and visual inventory.
• Mining — LiDAR for stockpile volumes and pit design compliance, orthomosaic for site overview and planning.
• Infrastructure design — LiDAR for terrain model feeding into civil design software, photogrammetry for context and presentation.

Dual-sensor drones

Several current drones carry both LiDAR and camera sensors:

Flying both sensors simultaneously (or sequentially on the same platform) reduces mobilisation costs and ensures the datasets are temporally aligned.

Processing comparison

Processing time is where LiDAR has a significant practical advantage on larger projects. A 500-hectare photogrammetry dataset with 15,000+ images can take days to process, even on a powerful workstation. The equivalent LiDAR dataset processes in hours.

Accuracy in practice

Published specifications don’t always reflect real-world results. Here’s what I typically see across projects:

The gap widens dramatically under vegetation. On an open construction site, the two technologies deliver comparable accuracy. Under a forest canopy, LiDAR is the only viable option.

Cost comparison

These are rough ranges and vary significantly by region, project size, and deliverable requirements. The key takeaway: LiDAR has higher upfront equipment costs but faster processing; photogrammetry has lower equipment costs but higher processing time.

For a survey company deciding where to invest, the question is volume. If you’re flying regularly and your clients need terrain data under vegetation, LiDAR pays for itself quickly. If you’re doing occasional small-site surveys with visual deliverables, photogrammetry is the practical choice.

Delivering the data

Regardless of which technology you use, the delivery challenge is the same: large, specialised files that your client probably can’t open without specialist software.

LiDAR point clouds in LAZ format, photogrammetry orthomosaics in GeoTIFF, 3D meshes in OBJ or GLB — these aren’t files that open with a double-click.

Browser-based platforms like Swyvl let you upload any of these formats and share them via a link. Your client opens the link, the data loads in their browser, and they can orbit, measure, and annotate without installing anything. It works for point clouds (via Potree), 3D models, orthomosaics, and GeoTIFFs.

For more on delivery workflows, see How to Deliver Drone Survey Data and File Sharing for Surveyors.

Summary

Neither technology is universally better. They solve different problems. The best survey companies understand both and deploy the right tool for each project — or combine them for the most complete result.