What is Photogrammetry

Photogrammetry is the science of making precise measurements of three-dimensional objects and terrain features from two-dimensional photographs. The input to photogrammetry is multiple photographs of some real-life object or a scene, taken from different angles and locations, and the output is typically a map, a drawing, a measurement, or a 3D model of the subject matter.

How does photogrammetry work?

This process works by using the method of triangulation. It involves taking pictures from a minimum of two different locations. These pictures create lines of sight that lead from each camera to specific points on the object being photographed. The intersection of these lines plays into mathematical calculations that help produce geometric 3D coordinates of the specified points, which sometimes is referred to as “point cloud”.

Photogrammetry triangulation creates 3D object

Photogrammetry history and evolution

Photogrammetry is nearly as old as photography itself. Since its development approximately 150 years ago, photogrammetry has moved from a purely analog, optical-mechanical technique to analytical methods based on computer-aided solutions of mathematical algorithms, and now to digital photogrammetry based on digital imagery and computer vision, without any mechanical hardware.

Topographic map creation was the first application of photogrammetry, which gave way to the development of simple map-making techniques geared towards the mapping needs of the military during World Wars I and II.

Types of Photogrammetry

Two general types of photogrammetry exist aerial (with the camera in the air) and terrestrial (with the camera handheld or on a tripod).

Terrestrial Photogrammetry

The photos used in photogrammetry can also be taken directly on the ground or from a fixed terrestrial position, such as by a person walking around or a camera mounted to a building.

Aerial (Drone) Photogrammetry

Aerial photogrammetry was introduced with the invention of flight, and drone technology advancements in the last 6 years have made it more accessible and affordable.

Now a drone captures several high-resolution photos over an area at different angles from multiple vantage points. These images are then processed with photogrammetry software (such as Pix4D, AgiSoft, Drone Deploy, etc.) that “stitches” the images together to create point clouds.

Photogrammetry accuracy

Several variables affect the overall accuracy of an aerial map, including camera specifications, the number of photos collected, the photo overlap percentage, the flight altitude, atmospheric conditions (wind speed, air pressure, air density, cloud cover, position of the sun), the GPS signal strength, and the ground sampling distance (GSD). A lower flight altitude yields a smaller GSD, and a smaller GSD yields a more accurate and higher resolution aerial map.

Many of these factors are already accounted for by main drone manufacturers that produce drones for collecting visual data used for photogrammetry processing. These drones are equipped with proper cameras, and the differences between those drones are oftentimes the quality of the camera itself, and the experience of the drone operator in using the drone for photogrammetry data collection (programming the flight path, calibrating, camera settings, etc.)

Relative and absolute photogrammetry accuracy

In aerial photogrammetry, there are such terms as relative (local) accuracy and absolute (global or spatial) accuracy.

Relative accuracy is the measurement of how objects are positioned relative to each other in a reconstructed model (Orthophoto map, Digital Surface Model, or 3D point cloud). Most aerial maps provide a high degree of relative accuracy. So, for simple use cases (measuring distances, volume, and height differences, etc.) relative accuracy should work just fine.

Absolute accuracy refers to the difference between the location of the objects on the reconstructed model and their true position on the Earth (or a geodetic coordinate system). Aerial maps do not provide a high degree of absolute accuracy unless Ground Control Points (GCPs) are used during the drone aerial data collection.

What are Ground Control Points (GCPs) in photogrammetry?

Ground control points are points on the ground with known ground coordinates, which can be easily recognizable in aerial photos. The ground coordinates can be measured with traditional surveying methods or an existing map (even Google Earth map), or smart control points that automatically upload position data, such as Aero Points. To calculate the coordinates for each point on the aerial photography several ground control points’ coordinates should be in place in the survey area and they should be clearly visible in the taken aerial photographs.

Drone Photogrammetry Outputs

The photogrammetry software generates object or area point clouds that enable data visualization, turning these clouds into various outputs for different applications, with many allowing for measurements such as distances, heights, areas, and volumes:

Orthomosaic maps are stitched from multiple drone photos that have been corrected for lens distortion, with each pixel containing 2D geo-information (X, Y) which enable the generation of measurements such as horizontal distances and surfaces.

Topographic maps show details of the relief features using contour lines (connecting points of equal elevation), giving a better understanding of the surface of the area.

3D Point Cloud is generated from drone images with each point containing geospatial (X, Y, Z) and color information. It provides a model for distance, area, and volume measurements.

3D Model is often generated from the 3D Point Cloud, and is a reproduction of the edges, faces, vertices, and texture of the object or area shot by the drone, and is most useful for visual inspection. This type of model can be used in CAD (Computer-Aided Design) or BIM (Building Information Modeling) software.

Digital Surface Model represents the elevation associated with the surface of the earth including topography and all human-made features, containing 2D information (X, Y) and the altitude (Z) of the highest point for the area.

Digital Terrain Model (also Digital Elevation Model) is a topographic model representing the bare ground surface without any objects like plants or buildings and can be uploaded into and manipulated by computer programs. The model contains 2D information of the area.

Digital Terrain Model. Credit: Jason Stoker

These outputs can be used in many applications in various industries, where the workflows for the collection of this type of data were largely manual and time-consuming. With drone photogrammetry productivity and efficiency of such operations can be improved significantly, and allow for more opportunities to innovate with such easy access to digitized data.

If you are interested in learning more about drone photogrammetry and its applications for your business, Contact us for a free consultation!