Photogrammetry

01Overview

Photogrammetry is an archaeological method that builds three-dimensional models of sites, structures, and objects from ordinary photographs. Instead of relying on specialized scanning hardware, it uses overlapping images taken from different positions and applies computational geometry to reconstruct the shape and surface of a subject. Because almost any digital camera can supply the raw images, the technique has become widely accessible and is used in field survey, excavation recording, museum documentation, and public outreach.

The central idea is that if the same point on an object appears in several photographs, and the camera positions are known or can be estimated, then the three-dimensional location of that point can be calculated by triangulation. Modern software automates this process, detecting matching features across images, estimating camera positions, and producing a dense cloud of points that describes the surface. The result can be textured, measured, and shared as a digital model.

Photogrammetry sits within the broader family of spatial recording approaches in archaeology. It complements methods such as lidar, aerial photography, and underwater survey techniques, and it is often combined with them. Its appeal lies in the balance it strikes between affordability, speed, and geometric detail, making it useful for both rapid field documentation and careful laboratory study of artifacts.

02What it is and how it works

At its core, photogrammetry is a set of mathematical procedures for recovering three-dimensional information from two-dimensional images. When a camera photographs a scene, each image is a projection of the three-dimensional world onto a flat sensor. If the same physical feature is visible in multiple images taken from different viewpoints, the differences between those projections encode depth. Photogrammetric software exploits this by identifying corresponding features and solving for the geometry that best explains all the observations.

The workflow typically begins with feature detection, where distinctive points such as corners, edges, and textured patches are located in each image. The software then matches these features between overlapping photographs. From the matches, it estimates the relative position and orientation of the camera for each shot, a process often called bundle adjustment. Once camera positions are known, the matched points can be intersected in three-dimensional space, producing a sparse point cloud.

A denser reconstruction follows, in which many more surface points are calculated to capture fine detail. This dense cloud can be converted into a mesh of triangles and draped with photographic texture to create a realistic model. Scale bars placed in the scene allow the model to be scaled correctly, so measurements taken from the digital result correspond to real-world distances. The entire process depends on good overlap, sharp images, and consistent lighting.

03How it is done step by step

The first step is planning and preparation. The archaeologist decides what will be recorded, whether a trench, a standing building, a small artifact, or an underwater feature. A camera is chosen, and settings such as aperture, shutter speed, and ISO are adjusted to keep images sharp and evenly exposed. Scale bars of known length are placed near the subject, and sometimes targets or markers are positioned to help the software align images. Consistent lighting is important, so diffuse light or a portable shade may be used.

The second step is image capture. The photographer moves around the subject, taking overlapping photographs from many angles. A common guideline is that each part of the surface should appear in at least three images, and often more, to ensure reliable matching. Images are taken in a systematic pattern, such as parallel passes or a circular orbit, keeping the subject centered and avoiding motion blur. For large sites, multiple sets of photographs may be needed, and for objects, a turntable or careful hand positioning can help.

The third step is processing. The images are imported into photogrammetry software, which detects and matches features, estimates camera positions, and builds a sparse point cloud. The user then generates a dense point cloud, cleans away noise, and creates a mesh. The mesh is scaled using the scale bars, and photographic texture is applied. Finally, the model is checked for accuracy, exported in a suitable format, and archived with metadata. The whole sequence can take minutes for a small object or hours for a complex site.

04What it reveals about the past

Photogrammetry reveals the three-dimensional form of archaeological evidence in a way that flat photographs and drawings cannot. It records the exact shape of a trench, the profile of a wall, the curvature of a vessel, or the surface detail of a carved stone. Because the model is measurable, researchers can revisit it to take distances, angles, and volumes long after the excavation has been backfilled or the object returned to storage. This makes it a powerful tool for documentation and for preserving a record of contexts that are inherently destructive to investigate.

Beyond measurement, photogrammetry supports interpretation. By viewing a model from any angle, archaeologists can examine how spaces were used, how structures related to one another, and how artifacts were positioned. Textured models can reveal subtle surface features such as tool marks, wear patterns, or inscriptions that might be difficult to see in the field. When combined with other datasets, such as lidar or aerial photography, photogrammetric models help build integrated understandings of landscapes and built environments.

The method also has a strong public and educational dimension. Digital models can be shared online, used in museum displays, or incorporated into virtual reality experiences. This allows students, researchers, and the public to explore archaeological subjects that are fragile, remote, or inaccessible. In this way, photogrammetry contributes not only to research but also to the wider communication and long-term preservation of archaeological knowledge.

05Strengths and limitations

One major strength of photogrammetry is its low cost relative to many other three-dimensional recording techniques. A basic digital camera and a computer with suitable software are often enough to produce useful models. The method is also flexible, working at scales from a single coin to a whole building, and it can be used in the field, in a laboratory, or underwater with appropriate housings. Because it is image-based, it produces visually rich results that are easy to interpret and share.

Another advantage is speed. Once images are captured, processing can be largely automated, and models can be generated quickly. This makes photogrammetry practical for rescue archaeology, where time is limited, and for creating records of features that will soon be removed. The resulting models can be measured, annotated, and compared, supporting detailed analysis and long-term archiving.

However, photogrammetry has limitations. It requires good overlap and sharp, well-exposed images; poor lighting, motion blur, or reflective, transparent, or featureless surfaces can cause errors or gaps. It reconstructs only what is visible in the photographs, so hidden areas are not captured. Accuracy depends on careful scaling and calibration, and large projects can demand significant computing power and storage. Finally, the method documents appearance and geometry, not material composition, so it is usually combined with other techniques for a fuller understanding.

06Common misconceptions

A frequent misconception is that photogrammetry is the same as laser scanning. Both produce three-dimensional data, but they work differently. Laser scanning measures distances directly by emitting light and timing its return, while photogrammetry infers geometry from images. Each has strengths: scanning can be very precise and works in some low-light conditions, while photogrammetry is often cheaper and captures photographic texture naturally. In practice, the two are sometimes used together.

Another misconception is that photogrammetry creates a perfect copy of an object. In reality, the model is an approximation built from a finite set of images. Its accuracy depends on image quality, overlap, scaling, and processing choices. Small errors can accumulate, and surfaces that were not photographed well may be incomplete or distorted. A model should therefore be treated as a measured representation, not as an exact replica, and its limitations should be documented.

Some people also assume that photogrammetry is only for experts with expensive equipment. While advanced workflows exist, the basic principles are accessible, and many projects use consumer cameras and free or low-cost software. That said, producing reliable, well-scaled models still requires care, training, and attention to method. The technique is a tool within archaeology, not a replacement for archaeological interpretation or context.

07Frequently asked questions

Q: Can photogrammetry be used underwater? A: Yes. With waterproof housings or specialized cameras, divers can photograph submerged sites and artifacts, and the same processing steps apply. Underwater photogrammetry is often used alongside other marine survey methods, though water clarity and lighting can affect results.

Q: How accurate are photogrammetric models? A: Accuracy varies with camera quality, image overlap, lighting, scale control, and software settings. With careful fieldwork and proper scale bars, models can be accurate enough for many recording and analysis tasks, but users should always assess and report error margins for their specific project.

Q: Do I need special equipment? A: A digital camera, a computer, and photogrammetry software are the essentials. Scale bars are important for measurement, and additional lighting or targets can improve results. Many accessible software options exist, making the method widely usable in archaeology.