How to 3D Survey Shrines and Temples｜A 6-Step Guide to the Flow of Investigation and Documentation

On-site investigations and documentation of shrines and temples increasingly demand accurate records of buildings and precincts. Roof warping and pillar tilt, details of carvings and bracket complexes, level differences of stone steps and approach paths, and relationships with surrounding topography—information that is hard to convey with drawings or photographs alone—can be preserved in a three-dimensional, objective form using 3D surveying methods.

Shrines and temples present particular challenges distinct from ordinary buildings. Many components have high historical value and require avoiding physical contact; it is necessary to reconcile capturing the overall precinct space with recording building details; consideration for worshippers and ceremonies is required; and a variety of materials—wood, roof tiles, metal, stone—coexist. Site conditions are therefore complex. Simply bringing equipment and measuring is unlikely to produce usable deliverables. It is important to design the whole workflow—from defining objectives, through on-site survey, acquisition, processing, to post-delivery use—as an integrated process.

This article organizes and explains the basic workflow for 3D surveying of shrines and temples in six steps for practitioners. It is compiled carefully to be useful both for those considering introduction for the first time and for those already familiar with photogrammetry or point-cloud acquisition, including common on-site pitfalls. By the time you finish reading, you should have a clear view of the order in which to proceed with 3D surveying of shrines and temples, where quality differences arise, and how to link deliverables to investigation and documentation tasks.

Table of Contents

• Why 3D surveying of shrines and temples is needed

• Step 1 Clarify the measurement objectives and deliverables

• Step 2 Survey site conditions and plan the measurement

• Step 3 Establish control points and georeferencing concepts

• Step 4 Acquire 3D data on site

• Step 5 Organize acquired data and create models

• Step 6 Package records into formats usable in practice

• Practical tips to avoid failure in 3D surveying of shrines and temples

• Conclusion

Why 3D surveying of shrines and temples is needed

Traditional investigation and documentation of shrines and temples has emphasized measured drawings, photographs, and written records. These remain indispensable methods, but when you want to grasp both the overall building shape and the condition of details at the same time, information tends to become fragmented. Plan and elevation drawings make three-dimensional relationships hard to see; photographs depend on shooting location and field of view and can become difficult to identify over time as to which part of the site was recorded.

This is where 3D surveying is effective. By acquiring buildings and terrain as data with three-dimensional position information, you can later inspect arbitrary cross-sections, recheck dimensions, and share the same geometry among stakeholders at remote locations. It is especially useful for pre- and post-renovation comparisons, recording aging changes, disaster recovery references, and preliminary considerations for preservation repairs.

Shrines and temples contain many elements that are difficult to organize in two dimensions: complex roof slopes, deep eaves, overlapping members, precinct undulations, and arrangement of stone features and lanterns. Often you need to understand the precinct as a whole together with the main hall, worship hall, gates, corridors, stone walls, and approach paths. Using 3D data as a foundation makes it easier to link individual records.

However, 3D surveying is not omnipotent. Even dense data can be unusable if the resolution is inappropriate for the purpose, and ambiguous georeferencing makes temporal comparisons difficult. Site-specific conditions—trees in the precinct, pedestrian traffic, backlighting, wet stone, narrow crawl spaces, or heights—can all affect quality. That is why it is necessary to clarify the meaning of the survey at the outset and proceed in a structured sequence.

In practice, it is important not to treat 3D surveying as merely a new technology. Clarify where and how it will be useful within the series of tasks—investigation, documentation, drawing production, maintenance management, and preparation of explanatory materials—and build acquisition methods according to those objectives. This is the quickest path to success. From the next section, we will go through concrete steps in six procedures.

Step 1 Clarify the measurement objectives and deliverables

The first step is to make clear why you are conducting 3D surveying. If this remains ambiguous, you may miss necessary areas on site or, conversely, collect an excess of high-density data that burdens processing. For shrines and temples, required accuracy, target range, and deliverable formats vary significantly depending on the purpose.

For example, if the goal is to understand the layout of the entire precinct, broad spatial relationships including approach paths, stone steps, and surrounding terrain are prioritized over highly detailed expression of individual buildings. Conversely, if the purpose is a pre-conservation inspection for preservation repairs, local shape accuracy—such as roof deflection, pillar tilt, dimensions around bracket complexes, and the condition of carvings—is more important. If the data will be used for future comparative records, you need georeferencing and management methods that make it easy to overlay subsequent measurements, not just data that looks good this time.

At this stage, first define the scope. Decide whether it is the entire precinct, only the main buildings, specific faces, or whether to include stone features and surrounding retaining walls. Next, consider the required level of detail. If capturing roof shape is the main objective, you do not need to capture the entire precinct at high resolution. On the other hand, if you must record decorative members, you should incorporate close-range photography or localized measurement tasks separate from broad-area acquisition.

Also envisage the final deliverables from the outset. Will you use point-cloud data as-is, organize a three-dimensional model, create cross-sections and elevations, or output images for a report? Necessary preprocessing and classification methods differ according to this. It is common for practitioners to be satisfied simply with having acquired data, and later fail to define how to use it, resulting in underutilization internally or externally.

There are many stakeholders at shrines and temples: preservation and repair planners, managers, contractors, researchers, and local community members, each wanting different information. Therefore, organizing the intended recipients of deliverables in advance is important. Not all recipients are familiar with specialized point-cloud viewers, so consider providing more accessible formats as needed. Planning in advance how to combine three-dimensional data with extracted plans, elevations, sections, and comparison diagrams prevents later deviations in downstream processes.

It is also effective to set priorities for measurement targets at this stage. On site, work may not proceed as planned due to weather, time, or access restrictions. Distinguishing mandatory targets from those to capture if time allows makes on-site decision-making easier. For example: exterior perimeters and plinths of main buildings are essential, surrounding trees are reference only, and detailed ornaments are conditional extras.

In short, the essence of Step 1 is not the measurement itself but designing deliverables that will be usable. Because shrines and temples are valuable and complex, clarity of purpose determines quality. Firmly establishing this step streamlines subsequent site surveys and data acquisition and increases the persuasive power of the deliverables.

Step 2 Survey site conditions and plan the measurement

Once objectives and deliverables are decided, the next step is to survey site conditions. Misreading site conditions is a major cause of quality degradation in 3D surveying of shrines and temples. In addition to complex building shapes, precincts are strongly affected by external environments. Tree branches and foliage, reflections from stone pavements, seasonal sun conditions, worshipper traffic, scheduled festivals and rites—there are many issues different from indoor building surveys.

During the site survey, first grasp the composition of the target. Check sightlines from which directions, the magnitude of elevation differences, obstacles, and areas you want to view closely. Deeply projecting eaves often create blind spots under eaves, and corridors or narrow passageways restrict acquisition positions. There may be insufficient space behind a shrine or visibility blocked by trees. Organize these conditions on plans and envision from where and in what sequence you will capture data.

Next, confirm access and work-time constraints. Shrines and temples are spaces associated with religious activities and worship, so available working hours may be limited. Verify opening hours, event schedules, peak crowd times, and permitted ranges for photography and equipment setup in advance, and reflect these in the measurement plan. Early morning may be preferable in some cases, while in others avoiding strong morning or evening oblique light and choosing daytime might be better. Decide the schedule considering both lighting conditions and usage patterns according to purpose and targets.

Also pay attention to material characteristics. Wood is easy to capture in shape but loses information in dark areas; roof tiles and metal parts can be affected by reflections. Stone appearance changes when wet; moss or fallen leaves can confuse surface recognition. Plain white walls or monotone surfaces lack features and can make image-based processing unstable for alignment. Understanding these conditions helps you plan necessary supplemental photography and measurement directions in advance.

Safety must also be considered during site surveys. There are many locations where carrying and moving equipment poses risks—stone steps, slopes, wet floors, confined spaces, and heights. Avoid contacting culturally valuable components, do not obstruct worshipper circulation, and avoid creating trip hazards with temporarily placed equipment. Especially when working with multiple people, deciding standing positions and transport routes beforehand reduces confusion on site.

At this stage, prepare a simple measurement plan document. It should summarize the target scope, acquisition order, assumed measurement positions, common blind spots, precautions, contact persons, and weather-related contingency conditions. It need not be complex, but should provide shared on-site decision criteria. While flexibility on site is necessary in shrine and temple surveys, having a pre-plan improves the precision of responses.

The site-condition survey may seem mundane, but it greatly influences final deliverables. Discovering missing data later and revisiting is a heavy burden in terms of permission adjustments and scheduling. Therefore, in Step 2, read the site carefully and identify difficult-to-capture areas in advance.

Step 3 Establish control points and georeferencing concepts

In 3D surveying of shrines and temples, organizing georeferencing is as important as acquiring clean data. When you want to link precinct-wide data with detailed building records, or compare measurements over time, ambiguous georeferencing can greatly diminish the value of the deliverables. Even visually coherent three-dimensional data are difficult to overlay with data from other times or workflows if coordinate consistency is lacking.

First consider the spatial range over which positional consistency is required. Does relative shape of an individual building suffice, or do you need to treat the precinct layout integrally? Do you need correspondence with surrounding terrain or existing drawings? This determines how to establish control. Surveys of shrines and temples often require both local precision and overall consistency, since relationships among shrine buildings, stone steps, approach paths, torii gates, stone walls, and site topography are typically important.

This requires planning the systematic placement of control points or known-check points. Actual installation and measurement methods vary with site conditions, but at minimum you should proceduralize which datasets will be linked by what means. Even when circling a building to capture it, ensure there are sufficiently recognizable common points or surfaces so that front and back captures do not drift separately. If you acquire precinct-wide and detailed building data in separate operations, design them to share common references.

If you anticipate re-surveying, do not treat alignment as a one-off. For maintenance and management of shrines and temples, comparisons over multiple years are often desired. If last time’s and this time’s control do not match, it becomes difficult to tell whether observed differences are real changes or reference shifts. Therefore it is valuable to establish georeferencing concepts from the initial recording and preserve them in related documentation.

Georeferencing design must also be appropriate for site conditions, not just theoretical. Placing a control point where pedestrian traffic is heavy is hard to maintain; spots covered by leaves or in deep shade are hard to see. Avoid contacting cultural properties or affecting the site’s appearance; decide placements considering visibility, preservation, and safety. Rather than creating an overly complex control network, it is more operable to set up a configuration that meets the required accuracy and reproducibility without unnecessary complication.

In practice, it is also important to hand over records about georeferencing to downstream processes. If no one documents which references were used, in what order things were aligned, and which ranges were used for error checks, processing staff or future users will struggle. Since staff involved in shrine and temple surveys may change over long periods, recording the measurement assumptions contributes to long-term asset value.

Step 3 lays the groundwork for placing data in the correct spatial context. It is not flashy, but when done well, integration of broad and detailed data, drawing production, comparative analysis, and maintenance management all proceed smoothly. For long-term stewardship targets like shrines and temples, georeferencing design cannot be neglected.

Step 4 Acquire 3D data on site

With planning and georeferencing sorted, it is time to acquire data on site. In this step, the quality of your preparation directly reflects in the deliverables. In 3D surveying of shrines and temples, balancing broad spatial understanding with detailed shape recording is a major theme. It is often more effective to separate acquisition granularity by target rather than trying to capture everything in one pass.

In the field, first capture the skeleton of the entire precinct. Data that convey building layouts, approach paths, stone steps, elevation differences, and positions of major structures make it easier to place later detailed records within the overall context. Prioritize comprehensive coverage of extent here and plan to supplement details in subsequent steps. If you focus only on details from the start, global connectivity may be insufficient and integration later becomes difficult.

Next, carefully capture the exteriors and elevations of main buildings. Plan routes to cover not only the front but also the sides, rear, under eaves, and around plinths—areas prone to blind spots. If the roof has a large overhang, the visible range changes greatly with standing position and viewing angle. Solely frontal captures often lack information on eaves undersides and bracket complexes, so complement with oblique directions and positions at different heights.

For areas needing detailed records, consider close-range acquisition as a separate operation. Carvings, metal fittings, joints, cracks, losses, and suspected deformations often lack sufficient information at the distances used for overall capture. While adding close-up data, keep some overlap with the whole-site dataset so the positional relationship is not lost.

During acquisition, frequently check data quality on site. Assuming you will notice issues later in processing can result in missing parts, insufficient overlap for alignment, or weak information due to poor lighting. Revisit and add captures as needed; revisits are often not easy with shrines and temples, so on-site verification and additional capture are indispensable.

Human and environmental influences are unavoidable on site. Worshipper traffic, wind-blown foliage, wet floor surfaces, and strong sunlight shadows all affect acquisition quality and post-processing. The important thing is not to wait for perfect conditions but to plan acquisition order assuming condition changes. For example: capture corridor sections before pedestrian traffic increases, acquire the west-facing side before light shifts, and secure alternate angles for areas affected by moving branches.

Record-keeping on site is also essential. Note which areas were captured, where there were concerns, and reasons for additional captures to facilitate post-processing decisions. On sites that generate large volumes of data, it is easy to lose track of intentions later. For targets with repeated similar features, such as those common in shrines and temples, managing names and locations is particularly important.

The key points of Step 4 are to treat overall, main parts, and details separately, verify quality on site, and capture comprehensively. Measurement may seem like a discrete on-site task, but in reality it is the most critical phase determining downstream success. Every on-site decision directly affects the trustworthiness of the deliverables.

Step 5 Organize acquired data and create models

After on-site acquisition, move to data organization and modeling. The important thing here is not simply merging collected data into one file but preparing it in a usable state according to the objectives. For shrines and temples, broad precinct data, building exteriors, and close-range detail data often coexist at different granularities; without a clear processing policy, the output can be heavy and hard to use.

First, organize the acquired data. Separate by target, acquisition date, and range; standardize naming conventions; and check for gaps or duplication. If you have mixed in additional data captured on site, record why it was added to make processing decisions easier. Shrines and temples have many building and part names and many places with similar appearances, so ambiguous folder names or sequential numbering can create confusion.

Next, perform alignment and integration. This is where the georeferencing concepts from Step 3 come into play. While processing, consider how to link broad and detailed datasets and which ranges to prioritize for consistency. Judging solely by visual continuity can produce local alignment while creating distortions at the whole-site level. Conversely, forcing global alignment can degrade detail accuracy. Decide which level of consistency to emphasize according to the purpose.

Then, clean unnecessary parts and remove noise. Remove elements irrelevant to the purpose such as worshippers, temporary structures, vehicles, branches moved by wind, and reflection artifacts. However, over-deleting can erase contextual site information, so consider keeping both a version that preserves site context for investigative records and a cleaned version for drawing production or explanatory use. Because shrines and temples often require both overall aesthetic understanding and component-shape analysis, managing deliverables by use case rather than forcing all information into a single product is more practical.

In the modeling phase, determine what to represent and to what extent. Sometimes using point clouds as-is is appropriate; other times converting to surfaces makes comparisons and explanations easier. Attempting to convert everything into highly detailed solid models can create excessive workload and make updates difficult. For ornate subjects like shrines and temples, a practical approach is to distinguish parts that require detailed expression from parts where an overall shape is sufficient, depending on the objective.

Quality checks are also essential in this phase. Verify whether required dimensions can be read, whether cross-sections are usable, whether important areas concentrate missing data, and whether there are unnatural discontinuities between precinct-wide data and building details. Do not complete processing based solely on the processor’s viewpoint; check from the perspectives of those who will use the deliverables. Investigators need to see cross-sections clearly, maintenance managers need to track deterioration areas, and explanatory material creators care about legibility.

Data organization and modeling are not mere technical tasks but information editing. Real value in 3D surveying of shrines and temples arises when on-site captured information is converted into usable knowledge. This requires structuring outputs with downstream use cases in mind, not only optimizing processing efficiency.

Step 6 Package records into formats usable in practice

The final step is to package the completed 3D data into formats usable in practice. “Practice” here includes archival of investigation records, repair planning, stakeholder briefings, future comparisons, disaster verification, and facility management. No matter how high-quality the data, insufficient viewing methods or organizational rules will render the deliverables unused in storage.

First, make the deliverable structure easy to understand. Organize by use—entire precinct, individual buildings, detailed records, cross-section-ready data, and explanatory images—so contents are clear at a glance. Basic metadata such as file names, target names, acquisition dates, georeferencing, target scope, and processing history greatly improve reusability. Records of shrines and temples are often stored for long periods, so ensure someone unfamiliar with the project can understand the data.

Next, provide secondary deliverables in addition to the three-dimensional data as needed. For example: extracted major sections and elevations, location maps of deterioration, diagrams showing precinct elevation differences, and baseline maps for pre-renovation comparisons. While few people can handle 3D data, many stakeholders prefer paper, images, or conventional drawings. Because shrine and temple investigations involve a variety of stakeholders, organize outputs with attention to who needs to see what.

Also prepare explanatory materials for reuse. Summarize which ranges were captured by which methods, where blind spots exist, what missing parts there are, which georeferencing was used for integration, and what to watch for in future comparisons. Leaving only three-dimensional data without its assumptions makes it difficult to extract full value years later.

Preservation of deliverables is another important perspective. If data are stored only on an individual’s device or in folder structures understandable only to one person, they become practically unusable during handover. Define shared storage locations, archive destinations, viewing permissions, and distinctions between original and processed data so the organization can handle them. When planning continued monitoring, structure files to make future comparisons easy.

Furthermore, do not limit the deliverables to a single use. Although 3D surveying of shrines and temples may start for repair planning, it can later be used for public relations, education, visitor explanations, or disaster-response materials. Preserve data in a re-editable form with practical use in mind to broaden future applications. At the same time, consider whether and how to publish data and restrict expressions according to the nature of the targets and necessary management rules.

The essence of Step 6 is to turn measurement outputs from archived objects into operational assets. 3D surveying of shrines and temples only becomes meaningful when the data are used. Preparing for who will use them, where they will be used, and how they will be linked next substantially increases the value of the records.

Practical tips to avoid failure in 3D surveying of shrines and temples

We have covered the six steps, but in practice a succession of small judgments determines the outcome. Below are common failure points in 3D surveying of shrines and temples.

First, collecting data that is excessive or insufficient for the purpose. High-density acquisition is not always correct. If your objective is precinct layout confirmation and you capture mainly details, processing and management burdens increase. Conversely, if you need detail to assess deterioration and only broad-area data are acquired, critical decision material will be lacking. Design acquisition granularity starting from what you want to preserve.

Second, insufficient imagination about blind spots. Shrines and temples have many hidden areas due to eave overhangs, underfloor spaces, elevation differences, trees, and stone walls. What appears visible on site may be missing in processing. Cultivate the habit of thinking in advance where unseen surfaces will arise. Simply circling the exterior can leave out eave undersides and lower rear areas.

Third, underestimating georeferencing. Proceeding with the mindset that visual coherence is enough often leads to an inability to link broad and detailed data or to compare repeat surveys. For sites likely to be managed over time, treat georeferencing as an asset from the start.

Fourth, postponing on-site verification. If you do not check captures immediately and only realize gaps later, revisiting is burdensome. Quality assurance should be part of on-site acquisition. For critical areas, perform quick checks before moving on.

Fifth, failing to consider the deliverable recipients. Many stakeholders are unfamiliar with point clouds and 3D models; if you do not plan how to explain and hand over outputs, valuable records will go unused. Preparing secondary deliverables, legends, and usage-based organization is as important as technical accuracy.

Sixth, lacking management for continued use. Records of shrines and temples can increase in value over years. Structure them not only for immediate use but for re-measurement years later, pre- and post-renovation comparisons, and disaster reference. Recording acquisition assumptions, control, scope, and missing-data information dramatically improves reusability.

Considering these points, 3D surveying of shrines and temples is not merely photography or measurement but the work of designing records. Understand the value of the target, read site conditions, set georeferencing, and compile with usage in mind. Carefully running through this sequence is the shortest route to avoiding failure.

Conclusion

How you 3D survey shrines and temples depends more on how you design investigation and documentation objectives than on equipment selection. In practice, first clarify what you will preserve and for what purpose, survey site conditions, establish georeferencing, then acquire data with awareness of both overall context and details, and finally package the outputs into usable forms. The six steps presented here form the basic workflow.

Shrines and temples are complex in form, have high historical value, and require consideration of their usage environment. Therefore, ad hoc measurements are insufficient; you should first solidify your recording approach. Clarifying how to balance precinct-wide understanding with building-detail recording, how to set georeferencing robust enough for re-surveying, and who will use the deliverables greatly increases the value of 3D surveying.

Also, investigation of shrines and temples often requires attention not only to individual buildings but to surrounding topography and spatial relationships. Treating three-dimensional data as the spatial foundation and linking it with positional information as needed will become increasingly important. In particular, for sites that want to improve positioning and reproducibility of survey locations on revisit, considering three-dimensional records together with positional information raises practical accuracy.

In that sense, if you want to streamline investigation and documentation of shrines and temples, it is worthwhile to look beyond 3D surveying itself and also consider means to support on-site positioning. For example, using an iPhone-mounted high-precision GNSS positioning device such as LRTK can make on-site position information more manageable and facilitate management of measurement points within the precinct and linking to surrounding information. If you aim to develop 3D surveying of shrines and temples from one-off records into a continuously usable survey foundation, consider incorporating such positional-information utilization.