What is cultural heritage photogrammetry? Five basics to know before introduction

Interest in photogrammetry, which reconstructs three-dimensional shapes from photographs, is growing in the fields of cultural heritage documentation, restoration planning, and public utilization. The background to this is that the shooting and processing environments, which previously required large investments, have become easier to handle, and the method has spread as a practical approach for cultural heritage documentation. In practice, technical guidance for cultural heritage treats photogrammetry as a practical method for cultural heritage recording and states that it can be applied across a wide range of scales from terrain and large sites to small artifacts.

At the same time, even if practitioners become interested in the term cultural heritage photogrammetry, they often have practical questions such as “How accurate is it, after all?”, “How is it different from photogrammetry (photo surveying)?”, “Can it really be used for restoration or reports?”, and “What should be prepared on site?”. In the cultural heritage field, it is not enough to simply create a visually appealing 3D model. It is necessary to consider the reusability of records, future comparability, the reliability of information as a basis for conservation and repair decisions, and data management assuming long-term preservation. International professional organizations that promote best practices in cultural heritage information recording and management also organize their guidance around the idea that the important thing is not appearance alone but how to arrange information that will endure for recording and management.

Therefore, when introducing cultural heritage photogrammetry, it is essential to define “for what purpose, at what accuracy, which deliverables, and up to what operation they will be retained” before considering “whether it can be photographed”. The required shooting conditions and how to produce deliverables vary greatly depending on whether the purpose is conservation repair, current-condition recording, public exhibition, or research. Basic understanding before introduction will directly affect success or failure, because it prevents producing large amounts of data that cannot be used later.

Table of Contents

• What is cultural heritage photogrammetry

• Basic 1 It is a technique to create measurable data, not just record photos

• Basic 2 Shooting plans must be changed according to the object

• Basic 3 Accuracy is determined by shooting conditions and reference settings

• Basic 4 Deliverables are not limited to 3D models

• Basic 5 Introduce with the view of preservation, utilization, and operation

• How to proceed without failing with cultural heritage photogrammetry

• Summary

What is cultural heritage photogrammetry

Cultural heritage photogrammetry is a method that analyzes the shape and positional relationships of an object from multiple overlapping photographs and generates outputs such as three-dimensional models, orthophotos, and data usable for drawing. In the cultural heritage field, it is applied to diverse objects such as buildings, stone structures, wall surfaces, remains, excavation areas, terrain, and landscapes; in recent practice it is common to include workflows that reconstruct shape from photo sets, known as SfM. Official guidance for cultural heritage also positions photogrammetry as a practical recording method and emphasizes that it can handle targets ranging from wide-area to small-scale.

What is important here is not to understand photogrammetry merely as “creating 3D images.” What cultural heritage requires is not only to reproduce the appearance of an object attractively, but to record shape, dimensions, position, deterioration, and surface conditions as objectively as possible so that they can withstand future comparison and re-examination. For example, the extent of stone loss, bulging of wall surfaces, traces of finishing, stratigraphic relationships in an excavation, and undulations of the ground surface can be difficult to share by photographs alone. However, if photogrammetric deliverables are properly produced, it becomes easier to change viewpoints for inspection, read dimensions, and compare with records from other times.

Also, cultural heritage photogrammetry does not make traditional manual measurements or standard photography unnecessary. Rather, it complements existing recording methods and, as needed, is combined with other surveying or point cloud measurements to increase the density and reusability of records. Recent review studies similarly share the idea that, given the diversity of targets and purposes in cultural heritage, image-based methods should be combined with other measurement techniques.

In other words, the decision to introduce cultural heritage photogrammetry should not be based on “because it seems modern” or “because it looks good in 3D.” It is important to judge whether, for the object and purpose you want to record, reliable spatial information can be preserved from photographs, and where that data will be useful in investigation, preservation, restoration, or public use. Simply adopting this perspective will greatly improve the quality of pre-introduction consideration.

Basic 1 It is a technique to create measurable data, not just record photos

The first point to understand before introduction is that photogrammetry photographs are not “illustrative photos” but “photos for analysis and measurement.” Ordinary record photos are mainly intended to convey the features or atmosphere of the object. But in photogrammetry, three-dimensional positions are estimated using features that appear commonly across multiple images, so overlap between photos, continuity of viewpoints, focus, exposure, and how the subject appears directly affect data quality. Guidance for cultural heritage repeatedly shows that success and accuracy are greatly influenced by understanding shooting and processing.

If this difference is not understood, similar mistakes often occur in the field. For example, some people assume that having a few neatly taken front-facing photos is enough to produce a 3D model. In reality, if information from side directions is lacking, the three-dimensional reconstruction will be unstable, and if surface texture is sparse it is difficult to find matching points. Conversely, even if images look modest, if the object is photographed in an unbroken sequence that surrounds it, has sufficient overlap, and each image is of stable quality, those images become suitable for analysis.

This concept is particularly important in the cultural heritage field because, even when the object is fixed, actual site conditions are highly diverse. Outdoor stone objects are affected by changing light and moving shadows, indoor statues and murals have problems with low illumination, reflections, and cramped spaces, and excavation sites require rapid recording of exposed faces; positions accessible for shooting may be limited on scaffolding or in restricted areas. Under such constraints, success depends on whether shooting can shift from merely increasing the number of photos to consciously creating the connections needed for 3D reconstruction.

Furthermore, the valuable output of photogrammetry is not only the completed 3D model. The set of photographs taken at the time of shooting themselves become primary materials that can be reprocessed, which is important. If better analysis methods or different processing settings become available in the future, systematic retention of the original photos may make recalculation possible. Therefore, from the start, photos should be handled as recording materials, including naming photos, recording shooting order, documenting the target range, shooting date, and shooting positions.

Basic 2 Shooting plans must be changed according to the object

The second basic point is that there is no universal shooting method for cultural heritage photogrammetry. Even within cultural heritage, the shapes, sizes, and surface characteristics differ among castle stone walls, wooden buildings, murals, stone Buddhas, tumuli, excavation areas, garden remains, and excavated artifacts. Official guidance organizes that while photogrammetry can be applied across scales from terrain and landscapes to small artifacts, the conditions required differ by scale.

For example, when recording a building exterior, you need both photographs that capture the overall shape and photographs that capture detailed decoration and damaged areas. Only photographing the whole makes it difficult to read details, and only photographing details breaks down the overall positional relationships. The same applies to stone objects and sculptures: rather than simply circling at a constant interval, you must intentionally include viewpoints for shaded parts, scooped-out areas, and joints with the base where the shape changes, otherwise areas around losses or holes may be reconstructed poorly.

On the other hand, for excavation areas and large remains, overhead viewpoints are important. When planar records are needed, images taken at a stable angle relative to the plane, not only oblique photos, are required. Moreover, in excavation sites the situation changes over time, so “we can take it again later” often does not apply. It is necessary to decide in advance which stages to record and what to preserve as excavation progresses. Low-altitude aerial photography and multi-viewpoint capture have become widely used in cultural heritage in recent years, but the important factor is viewpoint design suited to the recording purpose, not the type of equipment.

Also remember that many cultural heritage targets are difficult to photograph. Glossy lacquer, metal, glass-protected exhibits, surfaces with repetitive patterns, uniformly white plaster, deeply shaded sculptures, and stone walls partially covered by vegetation all increase analysis difficulty. For such subjects, it may be necessary to adjust shooting times, equalize lighting conditions, remove unwanted items, or in some cases segment the subject for processing. The success of cultural heritage photogrammetry depends less on software performance and more on how well the field team reads target characteristics and incorporates them into a shooting plan.

Therefore, at the consideration stage, decide “what to photograph and how” before “with what to photograph.” For outdoor remains, consider weather and sun altitude; for building interiors, lighting and movement flow; for movable artifacts, whether rotation is possible and background conditions—adjust the checklist for each target. Equipment selection can follow. If the order is reversed, operations are likely to adapt unnaturally to the equipment and crucial recording quality may suffer.

Basic 3 Accuracy is determined by shooting conditions and reference settings

The third basic point is that photogrammetry accuracy is not automatically guaranteed. The ability to reconstruct three-dimensional shape from photos is not the same as ensuring accuracy sufficient for restoration planning or drawing. In surveying specifications for cultural heritage, spatial data including photogrammetry are emphasized as being “fit for purpose.” Where specific spatial accuracy is required, the surveying specifications, control points, control measures, and deliverable conditions must be designed.

Accuracy here has several layers. First, there is geometric stability—whether the shape of the object is reproduced without visible anomalies. Next, there is dimensional accuracy—how closely any measured dimension matches the real object. Furthermore, there is positional accuracy—whether data taken on different days or other survey results can be overlaid within the same coordinate system and reference. In cultural heritage practice, visualization for report publication may be sufficient in some cases, while for deterioration comparisons, consideration of repair extents, drawing production, or long-term monitoring, stricter positional and dimensional consistency may be required. If the purpose is not decided before shooting, none of these tends to be achieved well.

What is especially easily overlooked is how references are set. If a model is produced without a clear basis for scale, it may look coherent but be weak as measurable data. If you plan to compare multiple time periods or integrate with other survey results, incorporate reference points, known dimensions, and coordinate referencing from the initial stage. Latest surveying specifications for cultural heritage treat photogrammetry as a formal spatial measurement method alongside other measurement techniques and organize it within the context of accurate 2D/3D spatial data, control networks, terrain models, and monitoring.

Accuracy also changes during post-processing. If you uncritically include unnecessary photos, blurred images, insufficient depth of field, motion blur, images with large exposure differences, or images where the subject moved, the calculations may run but the results will be unstable. Conversely, if you clearly delineate the target area, use only high-quality images, and add necessary photos, you can achieve practically useful results without relying on excessively expensive environments.

What is really important in cultural heritage practice is not to insist on “X millimeter accuracy (mm (in))” but to clarify “which decisions this deliverable may be used for.” For example, the conditions required for capturing the current state of a wide-area site and for capturing fine deterioration on stone surfaces are different. If you decide before introduction whether the purpose is for viewing, drafting, restoration design, or temporal comparison, the necessary accuracy and site preparation naturally become clear. Accuracy is not something to be wished for after shooting; it is created by working backward from purpose and specifications.

Basic 4 Deliverables are not limited to 3D models

The fourth basic point is not to limit photogrammetry deliverables to three-dimensional models. Generally, photogrammetry is received as “3D model creation,” but in practice it is more important what kinds of deliverables can be derived beyond that. Official guidance notes that photogrammetric outputs are diverse and should be chosen according to analysis and presentation purposes. Orthophotos, 2D drawing basemaps, 3D geometry, and base data for comparisons are examples of differing outputs appropriate to various uses.

For instance, when documenting conditions on wall surfaces, floors, or stone facings, orthophotos that suppress viewpoint distortion are extremely effective. Ordinary photos taken on site may be easy to view, but because positional relationships and dimensions are not consistent, it can be difficult to accurately annotate repair extents or damage distribution. Orthophotos, by contrast, work well with drawings and record ledgers and make it easier to share the positions of cracks, losses, discoloration, and repair marks. For excavation areas, they serve as a baseline that is easy to compare later when arranging excavation surfaces or remains by time series.

Also, a 3D model becomes more valuable when used for cross-sectional checks, distance measurements, shape comparison, deformation assessment, and creation of explanatory materials rather than as a standalone visualization. When cultural heritage staff, designers, conservation specialists, contractors, administrators, and exhibition staff share the same object, it is effective for visualizing spatial relationships that are hard to convey by words alone and aligning the basis for discussion. Recent research summarizes that photogrammetry in cultural heritage has wide-ranging applications including historical records, geometric capture, public utilization, tourism use, and research support.

An often-overlooked point is the value of deliverables as comparative materials for the future. Even if you do not use them immediately for repair design, properly preserving current 3D records and orthophotos allows you to check changes after several years or compare conditions after disasters. Since cultural heritage is often irreversibly damaged, thinking “we don’t need detailed data now” is risky. Considering the minimum resolution and structure to preserve with future usability in mind raises the overall quality of documentation.

In short, the question before introduction should not be “whether to make a 3D model” but “which deliverables will help whose decisions.” If you specify uses like figures for reports, pre- and post-repair comparisons, explanatory meeting materials, public exhibition content, or research shape comparisons, shooting plans, processing policies, and data retention policies are less likely to wobble.

Basic 5 Introduce with the view of preservation, utilization, and operation

The fifth basic point is not to finish at shooting. Because shooting and analysis are conspicuous, attention in introduction discussions tends to focus on equipment and processing time. However, what truly makes a difference in cultural heritage practice is the design of subsequent preservation, sharing, updating, and reuse. Latest surveying specifications for cultural heritage clearly indicate that spatial measurement outputs including photogrammetry can be used for conservation plans, repair design, monitoring, long-term asset management, and digital archives. In other words, cultural heritage photogrammetry should be considered as a basis for long-term operation rather than a one-off event.

What becomes important here is the unit of data organization. If original photos, processing data, final deliverables, reference information used, shooting dates, target ranges, responsible persons, and processing conditions are not at least linked, reuse becomes difficult after a few months. Cultural heritage surveys often span fiscal years, and changes in staff or contractors are not uncommon. If the meaning of data is lost at each change, valuable 3D records cannot be utilized. That is why professional organizations that promote best practices in cultural heritage information emphasize information management as well as the records themselves.

In operation design, “who will view it” also matters. High-resolution data for researchers and lightweight viewing data for administrative or community explanations require different formats and ease of use. Even if you retain only high-resolution source data, it will not be used on site if viewing environments are not provided. Conversely, lightweight viewing data alone will not withstand future detailed analysis. To maximize the value of cultural heritage photogrammetry, design data by role—archival, analytical, and sharing—so that each purpose is served.

Furthermore, if you plan continuous use, consider comparability. If the initial shooting is done with special settings and later sessions vary, comparisons of temporal change will be difficult. Decide rules from the start for repeat use: shooting height, circulation method, reference placement, naming conventions, folder structure, and types of deliverables. With these rules, multi-year records become living assets. Because cultural heritage is a long-term object of care, it is necessary at the introduction stage to consider “can we preserve it next time in the same way?”.

How to proceed without failing with cultural heritage photogrammetry

Based on the five basics above, the key to succeeding with cultural heritage photogrammetry is pre-introduction organization rather than the technology itself. The first thing to do is clarify five points: target, purpose, required accuracy, required deliverables, and operation method. If these are decided, shooting plans, reference settings, and processing policies are less likely to wobble. If they are vague, the field tends to adopt a “take more photos just in case” or “just make a 3D model for now” approach, resulting in deliverables that are hard to use.

The next important step is pre-checking site conditions. Cultural heritage has more constraints than general survey targets, and elements affecting shooting include access permissions, lighting conditions, opening hours, conservation environments, scaffolding, contact restrictions, reflections and shadows, and surrounding obstacles. Although photogrammetry has become widespread in cultural heritage because it is more accessible and lower cost, misreading site conditions can still significantly degrade results. It should be understood that improved usability does not guarantee unconditional success.

When contracting or outsourcing, explicitly verbalizing deliverable specifications is essential. A vague request like “a set of 3D models” leads to differing assumptions by providers. What is needed is a clear statement of the coverage, intended uses, delivery formats, and the degree of comparison or measurement the data should support. Official cultural heritage specifications provide not only the specification body but also a project brief template because verbalizing purpose and specifications affects outcome quality.

In practice, do not aim to introduce the entire range perfectly from the start. Begin small with a clearly defined scope—exterior record, one face of a stone wall, one excavation unit, or a single stone object—and verify shooting procedures, data organization, and how deliverables are used to reduce failures. Based on those results, decide which targets to expand to next, where to reinforce positional control, and at which stage to combine with other survey methods; this makes introduction realistic and sustainable.

Cultural heritage photogrammetry is a technology that organizes the approach to recording more than its visual novelty. For this reason, successful sites attend carefully not only to shooting techniques but to record design, specification design, and preservation design. Whether you can adopt this perspective before introduction greatly affects the value obtained from the same photographs.

Summary

Cultural heritage photogrammetry is a technique that reconstructs three-dimensional shape using photographs and, at the same time, a way of thinking about spatial records to pass cultural heritage to the future. The basics to know before introduction are: take measurable photos rather than record photos, change shooting plans according to the target, understand that accuracy is determined by purpose and reference settings, recognize that deliverables are not limited to 3D models, and design with preservation, utilization, and operation in mind. Understanding these points will make it easier to generate data that is useful not only for one-off visualization but also for investigation, restoration, public use, and future comparison.

In particular, when recording cultural heritage including ruins, stone objects, buildings, and surrounding terrain, the handling of positional information as well as photographs can become important. The clearer it is where a photo was taken and what range was captured under what reference, the easier later comparisons and sharing become. If you want to proceed with current-condition recording in a more practical manner, organizing not only a photogrammetry shooting plan but also on-site reference creation and simple surveying concepts will make operations easier. In such cases, using iPhone-mounted high-precision GNSS positioning devices such as LRTK can help link positional information acquisition around cultural heritage with photographic records, making it easier to structure the recording workflow from site to office. Rather than ending with simply taking photos, integrating photogrammetry and positional information design to preserve usable cultural heritage records is becoming increasingly important in practical work.