BIM IQ

Photogrammetry vs LiDAR: Which Should Your Project Use?

By Kush Agarwal, President and CEO, YelloSKYE. Published 2026-10-26.

Executive Summary

• The Question: When is drone photogrammetry good enough, and when does a project need aerial or terrestrial LiDAR instead?
• The Answer: Choose photogrammetry for open, textured sites where a sharp map and terrain model are the main outputs. Choose aerial LiDAR when vegetation hides the ground or thin features matter, and terrestrial scanning when building elements must be measured precisely. Many projects use more than one, on shared control.
• The YelloSKYE Angle: BIM IQ and Survey IQ take photogrammetry and laser data on the same coordinates, so each part of a site can be captured with the method that suits it.

Photogrammetry or LiDAR is a choice between calculating geometry from photographs and measuring it directly with laser pulses. Photogrammetry is cheaper, produces a sharp colour map and is accurate over open ground with good control. LiDAR costs more but measures through gaps in vegetation, in poor light and on thin objects. The right pick depends on what is on the site and which decision the data has to support.

At YelloSKYE, both methods feed the same record: BIM IQ models buildings from laser and drone point clouds, and Survey IQ produces terrain data from drone surveys. This guide is a decision aid. It compares the two methods on the factors that usually decide the matter, then works through common construction and infrastructure use cases and suggests a choice for each. How the scanners themselves work is covered in laser scanning in construction.

The difference in one paragraph

A camera records light reflected from surfaces, and photogrammetry software infers depth by finding the same feature in several photos taken from different places. A LiDAR sensor fires its own laser pulses, times their return and records a distance for each one, so depth is measured rather than inferred. Everything else in this comparison follows from that: the camera needs light and texture and gives you colour; the laser brings its own light, ignores texture and gives you shape. How photogrammetry builds a model from photographs is explained in our guide to drone photogrammetry.

Aerial or terrestrial: two kinds of LiDAR

"LiDAR" covers two quite different jobs on a project, and mixing them up leads to the wrong purchase.

• Aerial LiDAR is a laser payload on a drone or aircraft. It scans the ground from above while a GNSS receiver and an inertial sensor track the platform's position and attitude. It is a terrain and corridor tool.
• Terrestrial LiDAR is a laser scanner on a tripod, or carried by hand or on a trolley, inside or around a structure. It is a building and element tool, used for as-built models, structural checks and plant rooms.

When people compare photogrammetry with LiDAR, they usually mean drone photogrammetry against drone LiDAR for site and terrain work. When the subject is a building, the real comparison is between terrestrial scanning and other ground methods, with drone capture covering the roof and surroundings.

Head to head on the factors that decide it

Two cautions on reading this table. First, accuracy figures quoted by sensor makers describe the instrument in test conditions; the accuracy of your dataset depends on control, flight or scan design and processing. Second, neither aerial method is precise enough for fabrication-level checks on a building, which is terrestrial scanning territory.

Vegetation: the deciding factor more often than any other

Photogrammetry can only rebuild what the cameras see. Over a closed canopy, scrub or tall crops, the surface model follows the top of the vegetation and the bare-earth model has to be guessed by filtering and interpolating. On a cleared site this hardly matters. On a wooded corridor or an overgrown parcel it can mean a terrain model that is wrong by more than the design tolerance in exactly the places a drainage engineer cares about.

A LiDAR pulse spreads slightly as it travels, and part of its energy can slip through gaps between leaves while the rest bounces off the canopy. The sensor records several returns from one pulse: the first from the treetops, later ones from branches and, often, the last from the ground. Classification software then separates ground returns from everything else. The result is a bare-earth surface that is measured, not estimated, beneath much of the cover.

Two practical points follow. Leaf-off seasons give LiDAR more gaps to work with in deciduous woodland, and very dense evergreen cover or thick undergrowth can still defeat it. And if the site will be cleared before work starts, a photogrammetry survey after clearance may serve the design better than a LiDAR survey before it.

Accuracy: what each method can prove

Over open, textured ground with surveyed ground control and checkpoints, drone photogrammetry produces terrain accurate enough for feasibility, design and earthwork measurement. YelloSKYE's drone surveys, flown with RTK-GNSS positioning and ground control points, achieve sub-5 cm horizontal and vertical accuracy. Aerial LiDAR reaches similar results on open ground and keeps them where photogrammetry cannot, under cover and on low-texture surfaces.

At building scale the picture changes. A terrestrial scanner measuring a column from a few metres away is far more precise than any aerial method, which is why as-built checks of structure and services are done from the ground. In YelloSKYE's as-built vs BIM clash detection work, the structure is scanned from the ground once a slab is poured or steel is up, and the cloud is laid over the design model to catch elements that are out of position while correction is still cheap.

Whichever method you choose, ask for the same evidence: errors on independent checkpoints for aerial work, and a registration report plus control checks for terrestrial scans.

Cost: where the money actually goes

The sensor is only part of the bill. A fair comparison looks at the whole job.

On an open, cleared site, LiDAR rarely repays its higher cost. On a long corridor through woodland, the arithmetic often reverses, because the alternative is a large amount of walked ground survey to fill the gaps photogrammetry leaves.

Outputs: the same names, different strengths

Both methods deliver point clouds, surface models and terrain models, but each is better at some of them.

• Orthomosaic: photogrammetry's strongest product. A LiDAR survey can carry imagery, but if the map itself is the deliverable, photogrammetry usually wins.
• Bare-earth terrain: LiDAR's strongest product wherever there is cover. On open ground the two are comparable.
• Point cloud: photogrammetry clouds are coloured and dense on textured surfaces; LiDAR clouds are cleaner on plain surfaces and capture thin features. What a point cloud is and how it is used is explained in our guide to what a point cloud is.
• Volumes: both work on open stockpiles and earthworks. Photogrammetry is the usual choice because it is cheaper to repeat, as in our earthwork volumetrics use case, where a highway project cut the wait for survey quantities from a fortnight to hours.
• As-built BIM: built from terrestrial scans for interiors and structure, with drone data for the envelope and site.

A decision table by use case

On linear projects, the choice can change along the route. Our linear corridor mapping work uses terrain-following drone flights to keep resolution steady along the alignment and maps corridors five times faster than conventional survey; a section through dense woodland is where a LiDAR pass would be added.

Using both on one project

The most efficient answer is often a mix: photogrammetry for the open site and the regular visual record, aerial LiDAR for the vegetated sections, and terrestrial scanning for the building once its frame is up. That only works if every dataset sits on the same survey control and height datum, so the clouds and surfaces line up without manual nudging. Setting up that shared control is the subject of our hybrid construction survey guide.

YelloSKYE's DEX platform holds drone photogrammetry, laser scans, 360° walks and fixed-camera feeds for a site on one timeline. On the modelling side, BIM IQ takes scans from the main scanner makers (Leica, FARO, Trimble, Matterport) alongside drone photogrammetry, reading .pcg, .pts, .xyz, .rcp, .las and .e57 files, and models them at a level of development between LOD 200 and LOD 400.

Questions to settle before you choose

• What is on the ground? If vegetation covers areas that matter to the design, plan for LiDAR there.
• Is the deliverable an image or a shape? A map for people to read favours photogrammetry; precise geometry of thin or plain features favours LiDAR.
• Is the subject terrain or a building? Terrain points to aerial methods; elements and services point to terrestrial scanning.
• How often will you repeat it? Frequent captures favour the cheaper method.
• What tolerance does the decision need, and what evidence will prove it was met?
• Will the site change first? Clearing or demolition can make a cheaper survey later the better choice.
• Are all datasets on one control network so the methods can be combined?

Explore BIM IQ