Photogrammetry uses photographs to derive information about the geometry of objects and surfaces. For a project team, its value is the ability to turn observations into a representation that can be inspected, shared and, when properly referenced and checked, used for measurements.
The result might be a point cloud, a textured model or a map. Understanding how those products relate to the photographs helps you decide what to request and what evidence should accompany the delivery.
The useful idea behind the technology
An individual photograph shows a projection of a scene. When the same visible feature appears in photographs taken from different positions, the relationship between those views provides information about its three-dimensional geometry. Overlap makes it possible to identify common features; suitable viewpoints help describe their shape.
Many digital workflows use Structure from Motion, usually shortened to SfM. The USGS definition of SfM describes the use of multiple views to derive geometric relationships. The processing depends on meaningful correspondences between images, rather than simply on having a large folder of photographs.
Consider a building photographed only from above. Its roof may be well represented while a wall has too little coverage. Adding more overhead images does not necessarily provide the missing view. The question is whether the surface of interest was actually observed well enough to reconstruct.
Choose the representation that answers your question
Photogrammetry names the measurement approach, not one mandatory output. A USGS technical report shows how overlapping imagery can support point clouds, elevation models and orthomosaics. These products serve different purposes.
If you need to inspect reconstructed spatial observations, a point cloud may be appropriate. If you want to communicate the shape and appearance of a structure, a textured mesh may be easier to view. Surface data can support height analysis, while an orthomosaic provides a corrected plan view of visible features.
An elevation model needs a clear description of the surface it represents. The tops of trees and buildings are different from the ground beneath them. A deliverable name alone does not explain how obscured areas were treated or which observations support the model.
Our guide to drone point clouds examines those distinctions in more detail. For example, a stockpile volume needs suitable surface information and a defined base; an attractive overhead image alone is not enough.
Separate visual realism from measurement quality
A photographic texture can make a model look convincing without proving its dimensions or position. Before relying on measurements, establish the required scale, coordinate references and quality checks. Evaluate the result against the job it is intended to support.
Image resolution and point count do not independently establish positional accuracy. The USGS guidance on drone imagery calibration explains why geometric control and validation matter. Ask for evidence of quality where the proposed use depends on a particular tolerance.
Missing observations also matter. Removing vegetation points cannot supply a measurement of ground that was never visible. Reflective surfaces, movement and limited visual texture can complicate reconstruction. Document those limitations so that a complete-looking display is not mistaken for complete information.
A drone is one way to capture the photographs
The technique is not limited to aerial imagery. The camera can be positioned on the ground or elsewhere, according to the scale of the subject and the views required. Choosing a capture platform is a separate decision from deciding to use photogrammetry.
Historic England's technical guidance discusses heritage recording from landscapes to small objects. That range illustrates the flexibility of the method. It does not imply that the same equipment settings or processing choices are suitable for every application.
Define the role of the data in your project
For construction, a survey can record visible conditions at a particular date. For mining, surface representations can support assessment of exposed areas and stockpiles. In heritage work, geometric records can support documentation and study of visible features. The intended decision determines the requirements in each case.
The model remains an input to professional analysis. It does not automatically establish structural condition, reveal concealed services or replace ground investigation. Useful questions for an initial discussion include:
- Which surfaces and features must be represented?
- Is the purpose visual communication, measurement or both?
- What reference information and quality evidence are needed?
- Which parts of the site might require additional observations?
Evaluate a practical use with Maply
The Maply platform provides drone image processing, map and model viewing, and measurement tools. Start an evaluation with a representative area and a real task, such as reviewing a surface or sharing survey observations with the project team.
When you contact Maply, bring the available imagery, expected outputs and the question those outputs need to answer. That connects the technology to a useful decision and makes the demonstration more informative than reviewing a model's appearance alone.
