Is SfM Effective for Cultural Heritage Preservation? 7 Use Cases and Cautions Explained

In the field of cultural heritage conservation and documentation, practitioners are required, within limited personnel and budgets, to preserve the condition of objects as accurately and as continuously as possible. In particular, buildings, stone structures, sculptures, wall surfaces, archaeological remains, and garden components deteriorate and deform over time, so it is important to preserve the state at a given point in time with high reproducibility.

In such situations, attention is focused on SfM. SfM is a technique for reconstructing three-dimensional shapes from photographs, and it is applied to a wide range of practical tasks such as the preservation, investigation, exhibition, and restoration planning of cultural heritage. However, simply taking photographs does not automatically produce high-precision three-dimensional records. The quality of the results can vary greatly depending on the material of the object, the shooting conditions, the required accuracy, and the operational framework.

This article is aimed at practitioners searching for information on "SfM cultural heritage" and organizes and explains why SfM is considered effective for cultural heritage preservation, seven concrete use cases, and the key points to keep in mind when introducing it. It is compiled to be useful not only for those considering new adoption but also for those already conducting documentation who are experiencing issues with the stability of results or the breadth of application.

Table of Contents

• What is SfM?

• Why SfM Is Attracting Attention in Cultural Heritage Preservation

• Use case 1: Recording and preserving the current state of cultural heritage in three dimensions

• Use Case 2: Comparing changes over time in deterioration and damage

• Use Case 3: Objectively organize the before-and-after condition of restoration

• Use Case 4: Expanding into exhibition displays and educational use

• Use Case 5: Utilizing damage records and recovery planning during disasters

• Use Case 6: Make it a shared reference document for survey planning and construction review

• Use Case 7: Continue accumulating as a regional archive

• Precautions when using SfM for cultural heritage preservation

• Summary

What is SfM?

SfM is a method that estimates camera positions and the shape of an object from multiple photographs to create three-dimensional models. In the cultural heritage field, it is used to produce three-dimensional records by photographing objects on site from multiple directions and reconstructing their shapes from that set of images. Conventional flat photographs are effective as visual records, but they have limitations when it comes to consistently capturing depth, surface relief, and spatial relationships. By using SfM, sets of two-dimensional photographs can be converted into outputs containing three-dimensional information, significantly changing how artifacts are preserved, compared, and shared.

In recording cultural heritage, it is important not only to preserve appearance but also to be able to verify later information that can inform practical decision-making, such as changes in shape, the extent of loss, surface deterioration, tilt, and displacement. SfM is well suited to those requirements, and if imaging is carried out appropriately, it can flexibly produce three-dimensional records of objects that are difficult to touch on site or that cover large areas.

However, SfM is fundamentally a reconstruction technique based on photographs. Consequently, it is heavily influenced by shooting conditions. For subjects with little surface texture or pattern, materials with strong reflections, low light conditions, backlighting, monotonous color surfaces, or complex shapes with densely packed details, reconstruction can become unstable. Because cultural heritage items have diverse materials and forms, considerations different from those used in general surveying or structural recording are necessary. In other words, SfM is effective for cultural heritage preservation, but it is important to understand that it is a technique that requires appropriate operational design as a prerequisite.

Why SfM Is Gaining Attention in Cultural Heritage Preservation

One reason SfM is attracting attention in the field of cultural heritage conservation is that it can readily combine high levels of documentation, reproducibility, and shareability. Cultural properties are difficult to return to their original condition once damage has progressed, and restoration decisions must therefore be made with great care. For that reason, it is important to preserve the current condition as faithfully as possible and to record it in a way that allows later verification. Because SfM is photo-based, it offers great flexibility for field application, with subjects ranging from small-scale sculptures to medium-scale building components and even parts of ruins.

Moreover, cultural heritage preservation is not a task that ends with a single survey. Regular inspections, before-and-after comparisons of restorations, accumulation of research results, explanations among stakeholders, and promoting residents’ understanding—many uses occur across a time axis. A major advantage of the three-dimensional data acquired by SfM is that it readily becomes an asset for future re‑verification and re‑comparison. If shooting conditions are standardized and scale and coordinate management are properly set up at the time of capture, it becomes easier to grasp changes several years later.

Moreover, in the field of cultural heritage there is always an information barrier of "you have to go to the site to understand." When preservation staff, researchers, designers, contractors, and administrative staff share the same subject, relying only on plans and a few photographs can easily lead to differences in understanding. Three-dimensional models obtained by SfM make it easier to share three-dimensional spatial relationships and the distribution of damage, and can serve as foundational documentation to support stakeholder consensus.

However, being in the spotlight does not mean that all cultural heritage documentation can be completed with SfM alone. Depending on the size of the object and the preservation objectives, it may be more reasonable to combine other measurement methods, existing drawings, visual inspection, and dimensional verification. The important thing is not to treat SfM as a万能 solution, but to determine where in the overall cultural heritage conservation process it can be integrated most effectively.

Use Case 1 Recording and Preserving the Current State of Cultural Properties in Three Dimensions

The most basic and highly effective use case is to record and preserve the current condition of cultural properties in three dimensions. Cultural properties undergo various gradual changes, such as weathering, cracking, loss, subsidence, discoloration, biological growth, and the expansion of repair traces. Small changes are easily overlooked, and often by the time they are noticed the deterioration has already progressed. For that reason, leaving a record that allows the state at a given point in time to be reviewed later is of great significance.

Records created using SfM contain more information than flat photographs and have the advantage of making it easier to grasp an object’s three-dimensionality, the continuity of surfaces, and even fine surface relief. For example, with stone monuments or stone Buddhas, shallow inscriptions and surface wear can sometimes be captured in three dimensions. For wooden decorations and elements of joinery, areas that tend to fall into shadow can be more readily understood through complementary multi-directional photographs. Even for portable items such as pottery and small artifacts, creating 3D records for pre-exhibit documentation and research materials increases opportunities to examine them without frequently handling the originals.

The value of records of the current condition can be hard to see at the time of the survey. However, when time passes and comparisons become necessary, that value increases dramatically. Being able to confirm previous conditions in three dimensions even after personnel changes is a major advantage for handover and long-term management. Continuity is important in cultural heritage conservation, and there is a need for recording systems that do not rely solely on personal memory or paper documents. SfM is one of the methods suited to building that foundation.

On the other hand, to make them useful as current-condition records it is necessary to establish rules for capture and management. If you do not organize and retain information such as the date of capture, coverage area, subject name, orientation, capture conditions, person responsible, and supplementary observations, it will be difficult to reuse the data later. Rather than being satisfied with producing only three-dimensional data, it is particularly important in the cultural heritage field to operate them as a record system that can be preserved over the long term.

Use Case 2: Comparing Changes Over Time in Deterioration and Damage

In cultural heritage conservation, it is important to understand when, where, and to what extent deterioration and damage have progressed. SfM is well suited to comparing these changes over time. For example, if the same subject is photographed regularly within the same range and the 3D results are accumulated, it becomes easier to detect the widening of cracks, the progression of surface flaking, loss of components, and signs of settlement or tilting.

Of course, precise comparisons depend on shooting conditions and the accuracy of alignment. Still, compared with using only two-dimensional photographs, a major advantage is the ability to grasp the spatial relationships of changes in three dimensions. In photo comparisons, even slight differences in shooting angle or distance can alter the appearance, making it difficult to determine whether something is an actual change or simply a difference in how the photo was taken. With SfM, because the subject can be viewed three-dimensionally, it is easier to interpret the distribution and continuity of deformations, and to explain them to stakeholders.

For objects such as stone cultural properties and outdoor archaeological remains that are susceptible to wind, rain, and changes in temperature and humidity, a system to continuously capture the progression of deterioration is indispensable. Merely magnifying and observing only specific locations can overlook overall deformation and relationships with the surroundings. Comparisons using SfM are effective because they enable observations that connect local details with the whole. Because changes in small chips or cracks can be related to destabilization of the entire structure, being able to view them from a wide perspective is of great importance.

However, if the purpose is to compare changes over time, the initial recording must be designed with future comparisons in mind. If records are made using different capture methods each time, the reliability of comparisons will decline. Standardizing the capture route, capture height, overlap rate, area covered, and auxiliary positioning methods can significantly affect the quality of ongoing surveys. In cultural heritage preservation, emphasis should be placed not only on the results of each individual session but also on whether the records form a continuous series.

Use Case 3 Objectively Organize the Conditions Before and After Restoration

SfM is also effective for objectively organizing the condition before and after restoration. In the restoration of cultural heritage, it is required to retain records in a form that can be shared by all stakeholders, showing what damage existed before restoration, how far intervention proceeded, and how the appearance and shape became after restoration. Restoration records are related to future restorations and evaluations, so they cannot be treated as mere work reports.

Recording with photographs alone is possible, but a three-dimensional model can more clearly indicate the location, extent, and changes in shape of repair areas. In particular, surface undulations and differences in thickness, the treatment of missing parts, and the connection conditions with surrounding areas can be difficult to convey with only two-dimensional images. Using SfM makes it easier to compare before and after restoration from nearly the same viewpoint, and makes the results easier to use as explanatory material.

Restoration is a highly specialized field in which differences in stakeholders’ perceptions easily arise. Conservation staff, technicians, researchers, administrative personnel, and local stakeholders each focus on different points depending on their standpoint. Three-dimensional outputs can readily provide a common foundation for such diverse stakeholders and are advantageous in terms of accountability. In cultural heritage restoration, it is important to be able to trace afterward what was changed and to what extent, and SfM can serve as a means to support that transparency.

Another advantage is that it also makes post-restoration monitoring easier. When it is necessary to observe subsequent stability rather than treating the restoration as the end, three-dimensional records taken immediately after restoration serve as reference materials. If the same area is damaged again in the future, this makes comparisons with the previous condition easier, thereby helping to improve the quality of preservation decisions.

Use Case 4: Deployment for Public Exhibitions and Educational Use

The purpose of cultural heritage preservation is not simply to protect it. It is also important to convey it appropriately, broaden understanding, and pass it on to the next generation. Three-dimensional outputs created with SfM have the advantage of being easily deployed for exhibitions and educational use. Even for cultural heritage objects that cannot be moved frequently or for subjects that have restrictions on public display, three-dimensional data make it easier to devise ways to clearly convey their shapes and details.

For example, three-dimensional representations can show parts that cannot normally be approached during a regular visit—ornamental features at height, the positional relationships of reverse sides and internal structures, and so on. Depth and surface relief that are difficult to convey in flat photographs become easier to understand with a three-dimensional model, aiding visitors’ and learners’ comprehension. In school education and community learning settings, they also make useful materials for conveying the value of cultural properties to people who cannot visit the site.

What is important here is that creating three-dimensional models for public use itself increases the secondary value of preservation records. If data acquired for preservation purposes can be repurposed for research, education, public relations, and exhibitions, the cost-effectiveness of the records and understanding within the organization are likely to improve. Cultural heritage conservation work can sometimes face the challenge of outcomes being hard to see, but SfM is one method that makes it easier for stakeholders to visualize the results.

However, when using data for public release, it is necessary to organize the presentation separately from the data management intended for preservation. If readability is prioritized too much, the original shape information and the assumptions about recording accuracy can become ambiguous. It is important to operate with distinct roles for research, preservation, and public release, and to make clear which outputs are intended for which purposes.

Use Case 5: Utilizing for Damage Recording and Recovery Planning during Disasters

Cultural properties are exposed to various disaster risks such as earthquakes, heavy rain, landslides, fires, and strong winds. After a disaster occurs, it is necessary to quickly assess the extent of damage and proceed with required emergency measures and recovery planning. SfM is also effective in this context. If the damage can be rendered three-dimensionally from photographs taken on site from multiple directions, it becomes easier to document collapses, displacements, scattered structural elements, and surface damage in three dimensions.

Immediately after a disaster, site access restrictions and safety concerns can make lengthy investigations difficult. Even in such situations, prioritizing the capture of photographs and reconstructing the scene in 3D during post-processing is an effective approach. Of course, this assumes safety is the top priority, but if you can photograph the entire subject in a short time, you make it easier for experts to examine the details later.

Also, if three-dimensional records from before a disaster exist, comparisons with the post-disaster condition become possible. It becomes easier to compare how much displacement occurred, what is missing, and which components were lost, helping to formulate restoration plans. In cultural heritage restoration, how well the original state can be understood is of great significance. For that reason, keeping records with SfM during peacetime also contributes to improved disaster response capability.

Furthermore, three-dimensional outputs are also effective as materials for explaining damage conditions. For internal administrative coordination, expert meetings, and community briefings, it is important that stakeholders share a common understanding of the site conditions. Collapse patterns and spatial relationships that are difficult to grasp from photographs alone become easier to understand when shown in three dimensions. In cultural heritage preservation, SfM is valuable not only as a record in peacetime but also as material that supports decision-making in emergencies.

Use Case 6: Use as a common document for survey planning and construction review

In practical work for the conservation of cultural properties, multiple matters—such as survey, restoration, temporary works, safety measures, transport in and out, and consideration of the surrounding environment—are progressed simultaneously. In this context, three-dimensional outputs created by SfM serve as shared reference materials for survey planning and construction review. Because the site’s shape and its relationship to the surroundings can be confirmed three-dimensionally, they are useful for considering where to start work, which areas should be protected, and how to plan scaffolding and work flow.

Cultural heritage sites often have more constraints than ordinary construction sites. Areas that must not be touched, parts that cannot bear loads, zones where visibility needs to be considered, and narrow circulation paths — delicate conditions accumulate. If information is shared only through floor plans and words, discrepancies in site understanding easily arise, but having three-dimensional deliverables makes spatial comprehension easier and improves the quality of coordination meetings.

Furthermore, understanding the condition of the subject in advance makes it easier to prioritize inspections and work. It also makes it easier to review which areas should be observed more closely, where damage is concentrated, and which recording angles are lacking, thereby contributing to improved accuracy in subsequent surveys. Using SfM not merely as a finished deliverable but as an information foundation that connects to future fieldwork is a key point in increasing its value.

Furthermore, in situations involving external contracting and collaborative research, its effectiveness as a shared reference document is considerable. Even when stakeholders are located remotely, it becomes easy to share the subject’s three-dimensional condition, making it easier to proceed with deliberations while reducing the number of on-site inspections. In cultural heritage preservation, reducing the on-site burden is also important, and in that respect SfM is an effective means of supporting coordination tasks.

Use Case 7 Continuously accumulate as a regional archive

The value of SfM is not something that is completed by a single project. Its value as a cultural heritage archive grows when data are continuously accumulated at the level of regions, facilities, and managing organizations. For example, if multiple stone monuments in the same area, parts of shrine and temple buildings, and elements that make up historical landscapes are documented successively, this will contribute not only to individual preservation but also to an overall understanding of the region’s cultural resources.

Cultural properties often have value not as isolated items but within their surrounding environment and historical context. In expanding preservation from point-based protection to area-wide documentation, the accumulation of three-dimensional records is useful. If the current condition is periodically recorded, it becomes an asset that allows past states to be referenced in future research and conservation planning. This serves as a foundation that supports continuous management even when personnel change or organizations are reorganized.

Also, adopting a regional archives perspective makes it easier to standardize rules for photography and documentation. By organizing how subjects are named, the scope of photography, methods for preserving outputs, linking to related photographs, and integration with management registers, you can create an operational recording system rather than one-off 3D digitization. In cultural heritage preservation, building a sustainable system is more important than introducing advanced technology.

SfM is a well-suited method as an entry point for building such a system. It allows relatively flexible capture workflows and can be easily adapted to different types of subjects. To preserve local cultural heritage over the long term, it is essential to adopt the perspective of gradually building up recorded assets that can be handed down to the future, rather than producing results that are only useful for the present.

Considerations when using SfM for cultural heritage preservation

SfM is effective for cultural heritage preservation, but there are also points that are easy to overlook when introducing it. First and foremost, being able to generate a three-dimensional model from photographs is not the same as producing results that can withstand conservation practice. Even if you obtain a visually pleasing model, the reliability of its scale may be low, important parts may be missing, or it may not be suitable for comparative use. It is important to first clarify the required accuracy and the necessary recorded content according to the purpose.

The first point to note is not to underestimate photographic planning. Low-overlap captures, extreme angle differences, backlighting, subject motion blur, shifting shadows, surface reflections, and missed coverage will greatly reduce reconstruction quality. Cultural heritage objects are often difficult to rephotograph, so on-site capture planning can determine the outcome. You need to decide in advance which parts of the subject to prioritize, whether to favor details or the whole, and how to reduce blind spots.

The second point is that suitability depends on the material and surface condition. Glossy surfaces, nearly transparent surfaces, monotonous surfaces with little pattern, and areas with many thin, delicate protrusions tend to make photo-based reconstruction unstable. Cultural properties include challenging subjects such as metal decorations, lacquered surfaces, worn stone surfaces, and wooden parts in dark areas. If you operate uniformly without understanding the target characteristics, there is a risk that necessary information will be omitted.

The third point is that management intended for comparison and long-term preservation is necessary. If files are simply created without a determined storage location, if the date of photography and information about the subject are not organized, or if the relationship to the original photographs is unclear, they will be difficult to use several years later. Cultural heritage records are meant to be preserved for the future. It is essential to adopt the approach of managing result data, original photographs, supplementary notes, and related registers as an integrated whole.

The fourth point is how to ensure dimensional verification and positional control. In cultural heritage conservation, there are cases where knowing the shape is sufficient and cases where a certain level of dimensional reliability is required. In situations where consistency of position and size is important—such as restoration design or displacement comparison—the challenge is how to secure auxiliary reference information instead of relying solely on photography. If this is left ambiguous, you may be able to reproduce the appearance but lack robustness as documentary evidence for decision-making.

The fifth point is to consider both the on-site burden and the post-processing burden. SfM offers high flexibility in shooting, but if you underestimate the effort and verification work required for post-processing, operations will not be sustainable. As the number of targets increases, tasks such as naming, organizing, quality checks, decisions on re-shooting, and comparison management also increase. To maintain continued use, an operational structure that does not rely solely on on-site personnel is necessary.

The sixth point is not to confuse public use with archival records. The elements to be prioritized can differ between three-dimensional outputs intended for presentation and record outputs intended for preservation. Data that have been reduced in size or optimized for readability for publication do not necessarily serve as substitutes for the original archival records. Separating the positioning of deliverables by intended use will stabilize operations in the long term.

And the seventh point is not to try to make SfM a standalone solution. Cultural heritage preservation is built from a combination of photographic records, on-site observation, dimensional verification, drawing organization, registry management, and location information management. SfM can serve as the core of this, but it only becomes useful in practice when connected with other information. The key to success is to design operations in line with the original objectives—preservation, comparison, sharing, and transmission—rather than making technology adoption itself the goal.

Summary

SfM is effective for the preservation of cultural properties. There are many practical advantages: it can document the current condition in three dimensions, be used to compare deterioration and damage, serve easily as explanatory material before and after restoration, and be readily deployed for exhibition display and educational use. Furthermore, when uses such as damage recording in disasters, common reference materials for survey planning and construction review, and the continued accumulation as regional archives are taken into account, SfM can be positioned not merely as a three-dimensional digitization technology but as an information infrastructure for cultural heritage preservation.

On the other hand, the quality of the results is heavily influenced by the imaging design, the understanding of the subject’s characteristics, the clarification of accuracy requirements, and whether a management system is in place. In the conservation of cultural heritage, it is more important that outputs can be compared later, used for decision-making, and preserved long-term than that they simply look good. When introducing SfM, it is essential to be clear about which cultural heritage you want to record, for what purpose, and to what extent, and to consider mechanisms that enable continued operation.

Also, to put three-dimensional records of cultural properties to practical use, it is important to link shape data with positional information. For example, when recording archaeological features, stone structures, and cultural resources scattered outdoors, whether you can accurately preserve where the data were acquired affects how easily subsequent re-surveys and comparisons can be performed. In situations where you want to efficiently verify site coordinates and manage recorded locations, using LRTK, an iPhone-mounted GNSS high-precision positioning device, makes it easier to establish positional reference standards for cultural property records. By combining three-dimensional preservation using SfM with high-precision position verification, cultural property records become more practical and easier to reuse.