Terrestrial laser scanning (TLS) is a ground-based LiDAR technique that uses a tripod-mounted scanner to capture millions of precise 3D points, building a detailed digital record of buildings, terrain, forests and infrastructure. Where airborne sensors look down from above, TLS looks across a scene from the ground, which is why it reveals what other methods miss: building facades, tunnel walls, cliff faces and individual tree stems.
For students, TLS is a clear window into how LiDAR works. For practitioners, it is a routine tool for surveying, monitoring and documentation. For researchers, it supports measurement that was once impossible without physical contact, such as estimating tree volume without felling a tree or tracking millimetre-scale change on a rock face.
This guide covers how TLS works, how accurate it really is, how to go from field scans to a georeferenced point cloud, and where the method performs best, with references to the research literature.
What Is Terrestrial Laser Scanning?
Terrestrial laser scanning (TLS) is a ground-based form of LiDAR (Light Detection and Ranging) in which a tripod-mounted scanner emits laser pulses and records the returns to build a dense 3D point cloud of its surroundings. It is also called ground-based LiDAR, tripod LiDAR, 3D laser scanning, high-definition surveying (HDS) or 3D reality capture. The Survey Association (TSA) in the UK lists several of these names for the same family of techniques.
TLS operates on the same physical principles as airborne LiDAR, but from a horizontal or near-horizontal viewpoint. That perspective is why it excels at buildings, bridges, rock faces, tunnels and forest stems, features that airborne sensors see poorly. For a broader introduction to the technology, see our overview of LiDAR in remote sensing.
This article goes beyond a definition. It covers measurement principles, error sources, equipment, the processing workflow, applications and open research questions.
[IMAGE PLACEHOLDER: Figure 1 · colourised point cloud of a building or street · your own screenshot · file: tls-colourised-point-cloud-building-facade.jpg]
Figure 1. A colourised TLS point cloud of a building facade. [Add scanner model, software and site.]
How Does Terrestrial Laser Scanning Work?
A TLS scanner is an active sensor: it emits laser light, measures the range and intensity of the reflected signal, and records the direction of each pulse to compute a 3D coordinate.
A rotating mirror deflects the beam vertically while the scanner head rotates horizontally. Many scanners cover roughly 300 degrees vertically and a full 360 degrees horizontally, so a single setup captures almost everything in line of sight. Each measurement combines a range with two angles, which the instrument converts to X, Y and Z coordinates in its own local system. Many instruments also take photographs during the scan so each point can be colourised, and the University of Otago’s Arras Tunnels project, for example, used a scanner that records colour and return intensity for every point.
Figure 2. A rotating mirror sweeps the beam through about 300° vertically while the scanner head rotates 360° horizontally. Each measured range and its two angles are converted to X, Y and Z coordinates.
Ranging principles
There are three ranging approaches you will meet in the literature and in manufacturer documentation:
- Time-of-flight (pulse-based). The scanner times a short pulse’s round trip and computes distance as half the travel time multiplied by the speed of light. These systems reach long ranges, often hundreds of metres or more, at lower point rates.
- Phase-shift. The scanner emits a continuous modulated beam and compares the phase of the outgoing and returning signal. This allows very high point rates, usually at shorter ranges.
- Waveform digitisation. The full returning signal is recorded rather than a single return, enabling better handling of partial hits such as vegetation edges. Doneus et al. (2009) discuss this for terrestrial scanning.
Surveyors sometimes compare TLS to a total station: it measures individual 3D points in the same way, but captures them at a rate of hundreds of thousands to millions per second rather than one at a time.
Figure 3. Time-of-flight scanners time a pulse’s round trip, while phase-shift scanners compare the phase of a continuous beam with its return.
What the sensor records
For each point the instrument records range and intensity, the strength of the returned signal, which depends on surface reflectivity and incidence angle. Educational material from the University of Southampton notes that some scanners operate in a single wavelength band, such as a 523 nm green laser, so TLS intensity is not a multispectral measurement.
How Accurate Is Terrestrial Laser Scanning?
Terrestrial laser scanners typically achieve ranging precision of a few millimetres at tens of metres, but the accuracy of a finished, registered and georeferenced dataset is usually in the millimetre-to-centimetre range, depending on the workflow.
PM Corp, a survey and scanning firm, quotes accuracy as fine as 3 mm at about 30 feet (roughly 9 m) for modern scanners. Treat any single figure with care: manufacturers state accuracy under specific conditions, and real projects add registration and georeferencing error on top. If you need to understand the difference between these terms, our guide to GIS data quality, error, accuracy and precision explains the concepts.
Main error sources
| Error source | What happens | How to reduce it |
| Instrumental | Range noise and angular error limit point precision | Use calibrated instruments; follow manufacturer warm-up guidance |
| Range and beam divergence | The laser footprint grows with distance. A beam divergence of 0.3 mrad gives a footprint of about 3 cm at 100 m | Scan closer or at higher resolution for fine detail |
| Incidence angle | Steeply angled surfaces give noisier, less reliable returns | Add scan positions to view surfaces more directly |
| Surface reflectivity | Very dark, shiny or wet surfaces alter return strength and range | Adjust settings; use targets or matte spray where allowed |
| Mixed pixels | At edges the footprint straddles two surfaces at different distances, producing false points between them. Southampton’s teaching material illustrates this with a log in front of the ground | Filter edge points; check geometry near depth discontinuities |
| Atmosphere and weather | Rain, fog, dust and temperature affect the signal | Avoid precipitation; record conditions |
| Registration and georeferencing | Misalignment between scans or to the coordinate system | Good overlap, well-distributed targets, reported RMSE |
Figure 4. When the laser footprint straddles the edge of a near object, signal returns from two surfaces at different distances, and the recorded range can produce false points between them.
Reporting uncertainty
Treat registration error and georeferencing error as part of your result. Report the registration RMSE (root mean square error) between scans, the residuals at control points, and ideally an independent check against surveyed reference points. Record the coordinate system and datum used, and see our explanation of datum in GIS if this is new to you.
What Equipment Is Used in Terrestrial Laser Scanning?
A TLS survey needs a scanner, a stable mount, a way to tie scans together and to the ground, and software.
- Scanner. The central unit containing the laser, mirror and (often) a camera.
- Tripod or mount. A stable platform reduces vibration and keeps the scanner levelled.
- Registration targets. Spheres or flat targets placed in the scene to link scan positions.
- Control. A total station or GNSS receiver measures target or control-point coordinates for georeferencing. The Otago project describes establishing a ground control network with GPS or a total station. See GNSS vs GPS for background.
- Built-in sensors. Many scanners carry an inclinometer and compass, and some integrate GNSS, but these provide only approximate orientation. Do not assume a TLS scanner is positioned by GPS and an IMU in the way a mobile or airborne system is.
- Power, storage and a field computer for control and quality checks.
- Software. Manufacturer packages control the scanner and often do on-site cleaning. Processing and analysis continue in software such as CloudCompare, vendor suites, or the packages in our list of GIS software applications.
Figure 5. A typical TLS kit: scanner, tripod, registration targets and control equipment. [Add your own equipment details.]
Scanner classes at a glance
Exact specifications change with each model, so check current datasheets before citing numbers. The table shows the general trade-off between scanner types.
| Class | Typical ranging method | Typical strengths | Typical use |
| Short to medium range, high speed | Phase-shift | Very fast capture, high point rates, compact | Interiors, facades, construction, heritage |
| Long range | Time-of-flight, often with waveform processing | Range of hundreds of metres to over a kilometre | Mining, cliffs, open terrain, forestry |
What Is the TLS Workflow from Field to Deliverable?
A TLS project runs through six stages: planning, acquisition, registration, georeferencing, cleaning, and analysis or modelling.
Figure 6. The six stages of a TLS project, with the output of each stage and the main error source to watch for.
- Plan the survey. Choose scanner positions so every target surface is visible from at least one setup, and so neighbouring scans overlap enough to register. Scan resolution should be set from the smallest feature you need to resolve. GIM International’s article on TLS stresses careful positioning to ensure sufficient overlap.
- Acquire the data. Set up on a levelled tripod, place targets, choose resolution and quality settings, and scan. Record weather, scanner positions and notes. Capture photographs if colour is needed.
- Register the scans. Registration brings all scans into one coordinate frame.
- Target-based registration matches the same targets in overlapping scans. It is reliable and easy to quality-check.
- Cloud-to-cloud registration aligns overlapping surfaces directly, usually with the Iterative Closest Point (ICP) algorithm (Besl & McKay, 1992). It needs less field setup but depends on enough overlapping, geometrically varied surface.
- Georeference. Transform the registered cloud into a real-world coordinate system using control points measured by total station or GNSS. Keep metadata on the coordinate system, instrument and processing so others can reuse the data.
- Clean and classify. Remove noise, mixed-pixel points, moving objects and unwanted surfaces, then classify points where needed (ground, vegetation, structure).
- Create deliverables. Typical outputs include meshes, cross-sections, 3D models, and surface models such as a DEM or TIN. For background on surface types, see TIN vs DEM.
[IMAGE PLACEHOLDER: Figure 7 · before-and-after screenshots of two scans misaligned, then registered (CloudCompare or similar) · your own screenshots · file: tls-point-cloud-registration-before-after.png]
Figure 7. Two scans before (left) and after (right) registration in [software]. [Add the registration RMSE.]
Always verify the result against independent reference points. Our post on ground truthing explains why remotely collected data should be checked in the field.
Where Is Terrestrial Laser Scanning Used? What the Research Shows
TLS is used wherever fine detail, vertical structure or hard-to-reach surfaces matter, including forestry, geomorphology, construction, mining and heritage documentation.
Forestry and vegetation structure
TLS can measure tree height, stem diameter, stem form and canopy structure without felling trees. Liang et al. (2016) reviewed TLS in forest inventories and described both its promise for plot-level measurement and the practical limits of coverage and occlusion. Raumonen et al. (2013) developed methods to build quantitative structure models (QSMs) of trees from TLS point clouds, which enable non-destructive estimates of woody volume and branching. Such outputs support biomass work that would otherwise require harvesting. For a wider context, read our post on remote sensing in forest management.
[IMAGE PLACEHOLDER: Figure 8 · TLS forest plot with individual stems, or a quantitative structure model of one tree · your own image or openly licensed dataset · file: tls-forest-plot-individual-stems.jpg]
Figure 8. A TLS plot in a forest, showing individual tree stems. [Add site, scanner and credit.]
Geomorphology and change detection
Repeated scans of cliffs, landslides and riverbanks make it possible to measure surface change over time. Lague et al. (2013) introduced the M3C2 method, which compares point clouds directly in 3D along local surface normals and reports a confidence interval for each change, so that real change can be separated from noise and surface roughness. This approach suits steep or complex surfaces where simple DEM differencing performs poorly. A TLS system was used in teaching material from Southampton to illustrate mapping the face of a sea cliff to detect weathering changes through time.
[IMAGE PLACEHOLDER: Figure 9 · cliff or slope scan with a colour-coded change map (red for loss, blue for gain) · your own screenshot or openly licensed dataset · file: tls-change-detection-cliff-m3c2.png]
Figure 9. Change between two TLS surveys of a cliff, coloured by distance (red = loss, blue = gain). [Add site, survey dates and method.]
Construction and infrastructure
TLS verifies as-built conditions against design, supports quality control and tracks deformation of bridges, tunnels and buildings. Rapid capture also reduces time on site and exposure to hazards, an advantage TSA highlights in its client guide. See applications of LiDAR in civil engineering for further examples.
Mining and quarrying
Scans of open pits, quarry faces and stockpiles support volume calculation, slope monitoring and safety assessment, particularly where access to the face is unsafe.
Heritage and archaeology
TLS records monuments, ruins and excavations in their exact spatial context. The Institute for Digital Cultural Heritage at LMU Munich uses laser scanning to document archaeological sites and combines it with other survey data to create digital reconstructions. The Otago Arras Tunnels project used TLS as its main technique for 3D modelling of the tunnels.
Figure 10. A TLS point cloud of a [monument, tunnel or bridge]. [Add site and credit.]
TLS vs Airborne LiDAR, Mobile LiDAR and UAV Photogrammetry
TLS gives the highest detail from fixed positions, airborne LiDAR covers the widest areas, mobile LiDAR covers corridors quickly, and UAV photogrammetry is a flexible, low-cost option for open surfaces.
Figure 11. The same site captured four ways. Viewpoint determines coverage and detail: TLS excels at vertical structures from fixed positions, airborne LiDAR covers regions, mobile LiDAR covers corridors, and UAV photogrammetry suits open surfaces.
| Method | Viewpoint | Relative detail | Coverage | Best suited to |
| TLS | Ground, fixed positions | Very high at short to medium range | Local (site, structure, plot) | Facades, tunnels, rock faces, tree stems, monitoring |
| Airborne LiDAR | Aerial | Moderate | Regional | Terrain, forests, large-area mapping |
| Mobile LiDAR | Vehicle or backpack | High | Corridors | Roads, railways, urban streets |
| UAV photogrammetry | Aerial, low altitude | High on textured surfaces | Local to medium | Open sites, orthomosaics, volumes |
Relative ratings are general and vary by equipment and project. For deeper comparisons, see photogrammetry vs LiDAR, LiDAR vs photogrammetry and LiDAR vs radar.
Choose TLS when you need millimetre-to-centimetre detail on a defined site, the target is vertical or occluded, or you must repeat measurements precisely over time. Choose another method when the area is large, access is limited, or surfaces are mostly horizontal and open.
What Are the Limitations of Terrestrial Laser Scanning?
The main limitations of TLS are occlusion, limited coverage per setup, data volume, cost and field logistics.
- Occlusion. TLS sees only what is in line of sight, so vegetation, objects and terrain create shadows. More setups reduce gaps but add time.
- Coverage. Each setup covers a limited area, so large sites need many scans and careful registration.
- Data volume. A single project can produce hundreds of millions of points, demanding storage and processing capacity.
- Surface and weather effects. Reflective, dark or wet surfaces and rain or fog degrade quality.
- Cost and expertise. Scanners, software and trained operators are significant investments.
Open research questions
Current research challenges include more automated and robust registration, machine-learning segmentation of point clouds, combining TLS with UAV and mobile data, and standardising how accuracy and uncertainty are reported so that studies can be compared.
Key Takeaways
- TLS is ground-based LiDAR that captures dense 3D point clouds from fixed tripod positions.
- Ranging precision is typically a few millimetres at short range, but total project accuracy depends on registration and georeferencing.
- Report registration RMSE, control-point residuals and independent checks, not just scanner specifications.
- TLS performs best on vertical, complex or hard-to-reach structures and for repeat monitoring.
- It is limited by occlusion, coverage per setup, data volume and cost, so match the method to the project.
How are you using terrestrial laser scanning in your own work? Share your applications and experience in the comments.
Frequently Asked Questions
How accurate is terrestrial laser scanning?
Modern scanners achieve ranging precision of a few millimetres at tens of metres. The accuracy of a finished dataset is usually millimetres to centimetres, because registration, georeferencing, range, surface type and weather all contribute error.
What is the difference between TLS and airborne LiDAR?
Both use laser ranging, but TLS scans from the ground and captures vertical and complex structures in fine detail over small areas. Airborne LiDAR scans from above and covers large areas at lower detail.
What is the range of a terrestrial laser scanner?
It varies by design. Phase-shift scanners typically work from tens to a few hundred metres at very high speed, while long-range time-of-flight scanners can reach hundreds of metres to over a kilometre. Check the datasheet for your model and for the reflectivity assumed.
What is point cloud registration?
Registration aligns scans from different positions into one coordinate system. It is done with shared targets or by matching overlapping surfaces, typically with the ICP algorithm, and the resulting RMSE indicates alignment quality.
Which software is used to process TLS data?
Manufacturer software handles control and initial processing, and open tools such as CloudCompare are widely used for registration checks, filtering and analysis. Deliverables then move into GIS or CAD environments.
Can TLS work in rain or darkness?
TLS works in darkness because the scanner supplies its own light, though colour is lost unless artificial light is added, as PM Corp notes. Rain, fog and heavy dust degrade the signal and are best avoided.