LiDAR vs. Photogrammetry: When to Use What

How each technology actually works

Terrestrial LiDAR scanners (Leica RTC360, Faro Focus and similar) spin a laser rangefinder through a full sphere, recording millions of individual distance measurements per second. Each point carries an XYZ coordinate derived directly from time-of-flight or phase-shift measurement, plus a reflectivity value. There is no interpretation step: the point is where the laser says it is.

Photogrammetry works differently. A drone or camera captures dozens or hundreds of overlapping photographs from different angles. Software then identifies common features across images and triangulates their 3D position, the same principle human stereo vision uses. The result is a dense point cloud or textured mesh, but every point is a computed estimate, not a direct measurement, which is why its accuracy is more sensitive to conditions on the day of capture.

Accuracy envelopes you can plan a budget around

Marketing accuracy numbers rarely survive a real jobsite. These are the ranges our Montréal crews commit to in writing, based on thousands of delivered projects across interiors, façades and open sites.

Vegetation, lighting and weather effects

Photogrammetry needs consistent, diffuse light and visual texture to triangulate reliably. Overcast Montréal afternoons in spring and fall are close to ideal; harsh midday sun creates blown-out highlights and hard shadows that confuse the matching algorithm. Snow cover is worse: a flat white field has almost no texture for the software to lock onto, which is one reason we schedule most drone photogrammetry work outside the deep winter months.

Vegetation is the other major failure point. A laser can often punt through gaps in tree canopy and still return a ground point, giving LiDAR a real advantage for topographic surveys of wooded or landscaped sites. Photogrammetry, by contrast, only sees what the camera sees. It reconstructs the top of the canopy convincingly but tells you very little about the ground underneath. If bare-earth elevation matters, for example on a site with mature trees near a proposed building footprint, LiDAR is the safer default.

Lighting matters less for LiDAR because it emits its own signal, which is why terrestrial and drone LiDAR both work in near-total darkness. This is a genuinely useful property for occupied buildings: we routinely scan hotel lobbies, office floors and retail units overnight, after occupants have left, without needing any supplemental lighting rig.

Interior work versus open sites

Interiors favor terrestrial LiDAR almost without exception. Confined spaces, mechanical rooms, ceiling plenums and stairwells are exactly the environment where a tripod-mounted scanner excels: it can capture a dense, occlusion-managed dataset from a handful of setup positions, and the resulting point cloud is accurate enough to model MEP routing at LOD 300 or higher.

Open sites, large parking areas, roofs, quarries and agricultural or forested land are where photogrammetry and drone LiDAR earn their keep. Coverage per flight hour is an order of magnitude higher than anything achievable on foot, and for orthomosaic imagery, site context renderings or approximate topography, photogrammetry's accuracy is more than sufficient at a fraction of the cost of drone LiDAR.

Most real projects are not purely one or the other. A typical downtown Montréal redevelopment might need terrestrial LiDAR for every interior floor, drone LiDAR for the roof and mechanical penthouse, and drone photogrammetry for a site-context orthomosaic used in the rezoning submission. Treating that as three separate vendor relationships is how projects lose weeks to coordination overhead.

Registering everything into one coordinate system

The technical step that makes mixed-method capture worth doing is registration: tying every dataset, terrestrial scans, drone LiDAR, drone photogrammetry, back to a common survey control network. We typically establish ground control points tied to NAD83 CSRS or a project-specific local grid before any capture begins, then reference every scan and flight to that network during processing.

Done properly, the client never sees three disconnected datasets. They receive one registered point cloud where the interior LiDAR of the lobby lines up with the drone LiDAR of the roofline and the photogrammetric mesh of the parking lot, all in the same coordinates, all traceable back to the same control network. Skipping this step is the single most common reason multi-method capture projects fail to deliver usable data.

Cost comparison

Cost per unit area is the variable clients ask about first, and it is genuinely different across methods. As a rule of thumb across our Montréal project history:

• Terrestrial LiDAR interior capture: highest cost per square foot, justified by the accuracy and detail needed for renovation and MEP work
• Drone LiDAR: mid-range cost, priced mostly by flight time and site access complexity rather than area
• Drone photogrammetry: lowest cost per acre for large open sites, but plan for reprocessing time if conditions were marginal on the day
• Hybrid projects: total cost is usually lower than running two separate full-scope vendor engagements, because mobilization and registration are shared

Decision matrix

When a client asks us to pick one method for a whole project, this is roughly the logic we apply.

Montréal-specific notes: winter and occupied buildings

Montréal's climate genuinely affects the choice. Snow cover from December through March removes most of the ground texture photogrammetry depends on and can bury the site features you actually need to survey, so we schedule photogrammetric site work for the snow-free months whenever the project timeline allows. LiDAR is far less weather-sensitive: interior scanning runs year-round, and even exterior drone LiDAR flights are viable in winter provided icing and wind conditions are within the aircraft's operating limits.

Occupied buildings, hotels, hospitals, downtown office towers with active tenants, are another local reality. A large share of our Montréal projects happen after hours or on weekends specifically because LiDAR does not require additional lighting and produces almost no noise, so a scan can proceed in a hotel corridor at 2am without waking a single guest. Photogrammetry in an occupied interior is rarely worth the trouble; the lighting requirements and the number of photographs needed to get usable coverage make it a poor fit compared to a tripod-mounted scanner.