Skip to content
Reality Capture· 12 min read·July 15, 2026

Terrestrial Laser Scanning vs Photogrammetry for Heritage: Which to Choose?

Terrestrial laser scanning and SfM photogrammetry both produce point clouds of heritage structures — but they behave very differently on complex carved surfaces, large dark interiors, reflective stone, and in Indian field conditions. This guide gives you the decision framework for choosing, combining, or switching between them.

Quick Answer

TLS (terrestrial laser scanning) is superior to photogrammetry for: dark or uniformly lit interiors where camera feature matching fails; large-volume spaces where scan geometry is more reliable than image overlap; and any surface where colour/texture is absent (plain limestone, unpainted concrete). Photogrammetry is superior for: colour-textured output; inaccessible elevations (via drone); large outdoor sites where scan setup time would be impractical; and budget-constrained projects. The most effective approach for major heritage documentation projects is a combined workflow — TLS for interior control and complex geometry, photogrammetry for exterior colour and inaccessible areas.

The first question clients ask when specifying a 3D heritage survey is usually: 'Should we do laser scanning or photogrammetry?' Practitioners who have used both know that this is the wrong question. The right question is: what are the surface characteristics, space dimensions, lighting conditions, accuracy requirements, and budget constraints of this specific project — and which combination of methods best satisfies all of them?

Both terrestrial laser scanning (TLS) and Structure-from-Motion photogrammetry (SfM) produce point clouds. But the mechanisms are completely different, the failure modes are completely different, and the conditions in which each performs best are essentially complementary rather than competing.

How Each Technology Works

TLS vs SfM photogrammetry: fundamental mechanism comparison

AspectTerrestrial Laser Scanning (TLS)SfM Photogrammetry
Measurement principleActive — scanner emits laser pulses; measures time-of-flight or phase shift to determine distancePassive — camera captures images; software finds matching features across images to calculate 3D positions by triangulation
Point cloud generationDirect — each laser pulse returns a 3D pointIndirect — dense cloud computed from aligned camera positions
Colour / textureRGB overlay from integrated camera (less rich than photography) or registered photographyFull photographic colour at full camera resolution
Lighting dependencyLow — active illumination; works in darkness; some difficulty with direct sunlight interferenceHigh — requires adequate and consistent lighting; dark spaces, high contrast, and moving shadows cause problems
Transparent/reflective surfacesProblematic — glass, mirrors, polished stone cause multiple returns or false readingsProblematic — glass, mirrors cause incorrect depth matching
Scale constraintSelf-scaling — range measurement is absolute; no external scale reference neededRequires scale reference (GCPs or known distances) for accurate scaling
Output colour richnessModerate — scanner cameras are good but not equivalent to DSLR photographyHigh — full photographic detail at camera resolution

Accuracy Comparison

Accuracy comparison by measurement type

Accuracy measureTLS (FARO Focus S350 or Leica RTC360)SfM photogrammetry (Metashape, controlled GCPs)
Single-point range accuracy (20m)±2–3mmNot a single-measurement method — point positions derived from triangulation
Surface point position accuracy±3–5mm at 20m range; degrades with distance±5–15mm depending on GSD, GCP distribution, image overlap
Registration accuracy (multi-scan)±3–8mm per registered scan pairN/A — photogrammetry does not register scans
Absolute accuracy (georeferenced)±5–15mm with total station tie-in; ±50mm with sphere targets only±10–30mm with quality GPS GCPs; ±5–15mm with total station GCPs
Surface detail resolutionPoint spacing: 3–6mm at 20m (adjustable)Point spacing: 2–10mm depending on GSD and image overlap — can be finer than TLS at close range
Deep recess / occlusion handlingPoor — laser cannot see behind objects; multiple scan positions requiredPoor — camera also cannot see behind objects; similar occlusion problem

Accuracy claims in scanner specifications are range, not real-world project accuracy

Scanner manufacturers quote 'distance accuracy of ±1mm' or similar. This is the single-pulse range accuracy under ideal conditions. Real-world project accuracy — after registration of multiple scans, with real surface roughness, real scan angles and real environmental conditions — is typically 3–10× worse than the specification figure. The same applies to photogrammetry software GCP residuals: the reported residual is not the positional accuracy of every point in the cloud.

Speed and Cost

Speed and cost comparison for heritage documentation projects

FactorTLSSfM Photogrammetry
Fieldwork speedSlower — each scan station takes 3–15 min; indoor spaces require many stations; typical output: 500–2,000m² per dayFaster — photography can cover 1,000–5,000m² per day; drone survey is very fast for outdoor areas
Setup per station3–10 minutes (levelling, target placement, scan acquisition)Near zero — photographer moves continuously
Equipment cost (buy)₹30–80 lakhs for a professional scanner (FARO Focus / Leica RTC360)₹1–10 lakhs for DSLR + drone setup capable of professional results
Equipment cost (hire)₹15,000–₹40,000 per day including operator₹3,000–₹10,000 per day for camera equipment; drone hire separate
Processing time1–3 days registration + processing for a major heritage building2–5 days for a comparable Metashape project with GCPs
Processing software costFARO Scene or Leica Cyclone — typically bundled with scanner; ₹0–₹5 lakhs for additional licencesMetashape Pro: ~₹40,000 perpetual; RealityCapture: pay per export
Skilled operator requirementHigh — scan registration requires significant skill; errors are not always visible without careful reviewModerate — GCP marking and optimisation require skill; image capture is more forgiving

Strengths and Weaknesses by Condition

Performance comparison by field condition

ConditionTLS performancePhotogrammetry performance
Dark interior (mandapam, enclosed shrine)Good — active illumination; operates in darknessPoor to impossible — feature matching fails in low light without high-quality flash setup
Bright outdoor carved surfaceGood — but may have some surface saturation on white polished stone in direct sunExcellent — rich colour texture; high surface detail at close range
Large open site (temple complex >1ha)Slow — many scan stations needed; registration error accumulatesExcellent for drone orthophoto; ground photogrammetry covers large areas quickly
Fine carved ornament (< 50cm scale)Limited — point spacing at close range is good but colour overlay is moderateExcellent — close-range photogrammetry can achieve sub-1mm point spacing with very high colour fidelity
Wet or damp surfaces (monsoon condition)Good — laser unaffected by surface moisturePoor — wet stone reduces feature-matching contrast; droplets cause point cloud noise
Uniformly coloured surface (white marble, plain lime plaster)Good — laser works regardless of texturePoor — SfM requires visual texture for feature matching; smooth uniform surfaces fail
Very long range (>50m ceiling, large vault)Good — scanner handles long range wellPossible with telephoto photography but reduced accuracy at very long range
Moving subjects (visitors, vegetation)Moderate — fast scanners minimise motion artefactsPoor — moving objects between images create noise in the dense cloud

Decision Table by Project Type

Recommended method by heritage project type

Project typeRecommended primary methodSupplementaryReason
Interior recording: dark mandapam, enclosed shrineTLSPhotography for texture overlayLighting conditions defeat photogrammetry; TLS works in darkness
Exterior facade: carved gopuram (accessible)Photogrammetry (ground-based)TLS for deep-recess details if budget permitsRich colour; high detail; cost-effective; GCP control for accuracy
Inaccessible exterior: upper tiers of tall gopuramDrone photogrammetryTLS for lower accessible sectionsOnly practical method for inaccessible heights; TLS for ground level detail
Large temple complex: site surveyDrone photogrammetry for exterior + site; TLS for key interiorsGNSS for site controlEach method for what it does best
Carved stone detail (< 2m², high accuracy)Close-range photogrammetryNone typically neededSub-1mm point spacing achievable; excellent colour
Smooth white marble interior (Mughal tomb)TLSPhotography for documentation only, not photogrammetrySmooth uniform surface defeats SfM feature matching
Pre-demolition emergency recordTLS (fastest comprehensive coverage)Photography for colour recordSpeed is critical; TLS coverage is faster for interiors
Budget-constrained project (NGO / INTACH)Photogrammetry onlyNone — cost constraintCamera equipment cost is a fraction of scanner hire

The Combined Workflow

For major heritage documentation projects — large temple complexes, significant colonial buildings, multi-structure heritage sites — the optimum approach combines TLS and photogrammetry, with each method deployed for the surfaces and conditions it handles best.

The integrated workflow runs as follows:

  1. 1Establish a total station control network across the site — this provides the coordinate framework that will register both TLS scans and GCPs for photogrammetry into a single consistent spatial system.
  2. 2TLS scan all interior spaces — all enclosed or dark spaces where photogrammetry will fail; survey difficult junctions and complex structural intersections where geometric accuracy is critical.
  3. 3Drone photogrammetry for all accessible exteriors — facades, rooflines, courtyards, surrounding landscape; place GCPs measured from the total station control network.
  4. 4Ground-level photogrammetry for carved details — close-range photography of important carved panels, mouldings and ornamental elements that require colour and sub-5mm detail; GCPs from total station.
  5. 5Register TLS scans and photogrammetry point clouds into the same coordinate system — using the total station control as the shared reference; merge into a single project in CloudCompare or the scanner's native software.
  6. 6Export unified point cloud for CAD/BIM production — the combined cloud gives complete coverage with each dataset contributing what it does best.

Total station control: the integration key

The ability to integrate TLS and photogrammetry point clouds into a single spatially consistent project depends on having a shared control network measured with a total station. GCPs for photogrammetry and scan sphere targets for TLS must all be measured in the same coordinate system from the same total station setup. Without this shared control, the two datasets will be georeferenced independently and may not align accurately when merged.

India Context: Availability and Constraints

Professional TLS equipment (FARO Focus, Leica RTC360) is available for hire from survey companies in all major Indian cities. In Tamil Nadu, Chennai-based firms hire scanners with operators; Coimbatore-based suppliers are emerging. Daily rates for scanner hire with operator are ₹15,000–₹40,000 depending on scanner model and distance.

The high capital cost of TLS equipment (₹30–80 lakhs) means that most Indian heritage practices rely on hire rather than ownership. For projects outside major metropolitan areas, equipment transport and technician travel add significantly to project cost.

Photogrammetry equipment — quality DSLR or mirrorless cameras (₹80,000–₹2 lakhs) and professional drones (₹1.5–₹5 lakhs) — is far more accessible at an ownership level, which is one reason photogrammetry dominates small and medium Indian heritage projects.

Processing for both methods is typically performed in the office, not on-site. TLS projects require FARO Scene, Leica Cyclone or equivalent registration software; Metashape handles photogrammetry. The combined workflow may use CloudCompare (free, open-source) for point cloud management and merging.

Common Mistakes

  • Choosing TLS for a large outdoor site because 'scanners are more accurate' — TLS scan station setup and registration for a large outdoor site is slower and more expensive than drone photogrammetry, which achieves comparable accuracy for most heritage survey purposes.
  • Attempting photogrammetry in dark temple interiors without adequate flash — the result is a partially-aligned model with significant holes in dark recessed areas; either use TLS for dark interiors or set up proper photographic lighting.
  • Not having a shared total station control network — registering TLS scans with sphere targets and photogrammetry with GPS GCPs independently, then assuming the two datasets will align — introduces systematic errors; always use total station for both.
  • Quoting only for one method without assessing site conditions — visiting the site before specifying the method is not optional for heritage surveys; the mix of dark interiors, carved stone exteriors and inaccessible heights that is typical of South Indian temples requires a method-by-condition specification.
  • Underestimating TLS registration time — registering a 50-scan project of a complex temple interior takes 1–2 full days in office; this is frequently omitted from project timelines and budgets.

Professional Practice

In professional practice, method selection is a site-visit-dependent decision, not a general policy. The specification should be: 'TLS for [listed areas]: mandapam, gopuram interior, sanctum approaches. Photogrammetry for [listed areas]: all four exterior gopuram facades, outer prakaram walls, courtyard at site level, carved panel details on accessible piers. Drone photogrammetry for [listed areas]: upper gopuram tiers and roof.' This level of specificity in the brief is what separates professional heritage survey from generic survey service.

Fee structures typically handle the combined approach by: separate day rates for TLS fieldwork (equipment hire + operator + project management), photogrammetry fieldwork (own equipment + operator), processing, registration and merging, and final deliverable production. Presenting this breakdown to clients helps them understand why a combined approach costs more than photogrammetry alone — and why the additional cost is justified for a complex site.

Key Takeaways

  • 1TLS and photogrammetry are complementary, not competing — each performs best in the conditions where the other struggles; dark interiors, smooth surfaces and complex structural geometry favour TLS; outdoor carved facades, large sites and colour-critical work favour photogrammetry.
  • 2The most effective approach for major South Indian heritage sites is a combined workflow: TLS for interiors and complex structural geometry, drone photogrammetry for inaccessible exterior elevations, ground photogrammetry for carved detail.
  • 3A shared total station control network — with both TLS targets and GCPs measured from the same network — is essential for integrating the two datasets into a single spatially consistent project.
  • 4Real-world project accuracy for both methods is typically 3–10× worse than the specification figure; claims of sub-millimetre accuracy for production heritage survey should be treated with scepticism unless supported by check point verification.
  • 5TLS equipment cost in India is 20–40× higher than comparable photogrammetry equipment; for budget-constrained projects, photogrammetry with well-controlled GCPs is the practical choice for exterior work.

Frequently Asked Questions

Is laser scanning more accurate than photogrammetry for heritage buildings?

In controlled conditions, TLS is more accurate at close range (±2–3mm single point vs ±5–15mm for photogrammetry). However, real-world project accuracy for both methods is governed by factors other than single-point accuracy: scan registration accuracy for TLS, and GCP quality and distribution for photogrammetry. For most heritage survey purposes at 1:50 drawing scale (where ±20mm is acceptable), both methods are adequate. TLS provides higher intrinsic accuracy for complex interior geometry; photogrammetry with total-station GCPs can achieve similar results for exterior work.

Can I do photogrammetry in a dark temple mandapam?

Only with proper photographic lighting — a set of LED panel lights or flash units providing uniform illumination across the space. Without artificial lighting, SfM feature matching fails in low-light conditions. Handheld flash (single point source) creates strong shadows that defeat depth matching at carving recesses. For dark interiors of any complexity, TLS is the practical choice; it operates regardless of ambient lighting conditions.

How much does TLS cost compared to photogrammetry for a typical temple documentation project?

For a typical South Indian temple complex of moderate size (1–2 hectares, 5–10 structures), an indicative comparison: TLS-dominant approach (scanner hire at ₹25,000/day × 5 field days + 2 processing days + software) ≈ ₹2–4 lakhs fieldwork cost. Photogrammetry-dominant approach (drone + camera equipment at ₹3,000–8,000/day × 3 field days + Metashape processing) ≈ ₹50,000–₹1.5 lakhs. Combined approach: ₹2.5–5 lakhs. The photogrammetry advantage narrows when large dark interior volumes require TLS coverage.

J

Jabendra Raja

Technical-Commercial Partner, Evergreen Origins

Jabendra Raja leads heritage survey practice at Evergreen Origins, with field experience using both FARO Focus and Leica RTC360 scanners alongside Metashape photogrammetry on Tamil Nadu heritage sites.