What is photogrammetry?
Introduction
Photogrammetry is the process of extracting measurements and creating 2D & 3D models from images. By capturing images from multiple angles photogrammetry allows us to create accurate representations of the real world. A digital camera is the key to capturing these images and any digital camera can be used, including ones mounted on drones which greatly increase the efficiency and capabilities of photogrammetric projects. Whether its large scale aerial surveys or detailed close range inspections photogrammetry is a versatile and reliable way to get accurate data. Photogrammetry uses pixel matching to measure distances which may not be as accurate as laser scanning.
Used to create orthomosaic maps, GIS layers and 3D models of the real world. There are two types: aerial photogrammetry and close range photogrammetry.
Overview
Photogrammetry is a powerful tool that turns photographic images into precise measurements and detailed 3D models. By capturing images from multiple angles photogrammetry allows us to create accurate representations of the real world. Its used in many fields including mapping, architecture, engineering and even film and game design. Whether its large scale aerial surveys or detailed close range inspections photogrammetry is a versatile and reliable way to get accurate data.
Definition and History
Photogrammetry is the science and technology of obtaining information from physical objects and the environment by recording, measuring and interpreting photographic images and patterns of electromagnetic radiant imagery and other phenomena. The term “photogrammetry” was coined by Albrecht Meydenbauerde in his 1867 article “Die Photometrographie”. Over the years photogrammetry has evolved and has incorporated digital imaging and computational techniques. Today it encompasses a wide range of applications from 3D measurements from 2D data to colour ranges and values representing physical properties.
How does photogrammetry work?
Fundamentals
Photogrammetry uses the principles of optics and projective geometry to turn photographic images into 2D & 3D models. The process starts with capturing digital images of the target object or area from multiple angles. These images are then processed through several well defined stages to create digital models. The 3D coordinates define the location of object points in space and the image coordinates define the location of these points on the photographic images. The exterior orientation of the camera defines its position and direction of view and the inner orientation defines the geometric parameters of the imaging process. By aligning and processing these images photogrammetry software can create highly accurate digital representations of the real world.
Behind the scenes
While photogrammetry outputs like orthomosaics and 3D models are easy to use, the process behind the scenes is extremely technical. Advanced workflows include:
Tie Point Generation: Overlapping images are analyzed to find common points which are used to align and stitch the images.
Orthorectification: Geometric corrections to remove perspective and terrain distortions to create a map accurate image.
DEM and DSM Creation: Elevation models are derived from stereo imagery to analyze terrain and surface features.
These processes make photogrammetry robust across industries.
Types of Photogrammetry
Photogrammetry can be divided into two main types: aerial and terrestrial.
Aerial Photogrammetry involves capturing images from airborne platforms like drones, airplanes or satellites. This is used for large scale mapping and surveying. Aerial photogrammetry can be further divided into:
Satellite Photogrammetry: Uses satellite images to create 3D models and maps for large areas like cities or countries.
Aerial Photography: Taking photos from aircraft, used for small scale projects like local mapping and surveying.
Terrestrial Photogrammetry involves capturing images from ground based cameras. This is used for detailed inspections and small scale projects. Terrestrial photogrammetry can be divided into:
Close-Range Photogrammetry: Taking photos from a short distance, less than 100m, used in architecture and engineering.
Long-Range Photogrammetry: Capturing images from more than 100m, used for large scale mapping and surveying.
Also various photogrammetry methods exist:
Stereophotogrammetry: Two or more cameras to create 3D models.
Multiview Photogrammetry: Multiple cameras to capture different angles for a full 3D model.
Structure from Motion (SfM) Photogrammetry: One camera to create 3D models by analyzing the camera motion.
Photogrammetry software is used to process and analyze the images. Popular software are Agisoft Metashape, Pix4D and Autodesk ReCap which offer various tools to create 3D models and maps. This technology is used across many industries so it’s an essential tool for professionals working with 3D data.
Aerial Photogrammetry
This involves capturing images from platforms like drones, airplanes or satellites. The sensor faces downwards (nadir imagery) to capture overlapping images along a flight path. Aerial photos are used in mapping and surveying as they provide visual representation from airborne devices. These images are processed into digital elevation models (DEMs) and orthomosaics. Ground sample distance (GSD) is crucial in mapping as it determines the scale and detail of the images.
Orthomosaics are geometrically corrected to ensure uniform scale so they can be used for measurements and mapping.
The photogrammetric process corrects for image distortions like perspective geometry and terrain displacement to meet the accuracy required for mapping and analysis.
Close-Range Photogrammetry with Overlapping Images
Unlike aerial methods, close-range photogrammetry is used for objects or structures like buildings, bridges or utility towers. The sensor orientation varies (horizontal, oblique or upward) depending on the target. Focal length is critical to minimize image distortion and ensure accuracy. This method produces detailed 3D models and orthoimages for engineering and architectural purposes, focused on surface features not landscapes.
Orthomosaic and Digital Terrain Models
A output of photogrammetry is the orthomosaic—a seamless image corrected for scale and terrain. This requires precise data inputs:
Ground control points (GCPs) for positioning.
Tie points from overlapping images.
Digital terrain models (DTMs) to account for elevation changes.
When a DEM is not available, digital surface models (DSMs) derived from stereo imagery are used in the orthorectification process. DSMs are created by capturing overlapping images from different angles to generate 3D point clouds through image matching.
For aerial photogrammetry, stereo imagery is used to create elevation models like DTMs (bare earth) or DSMs (including buildings and vegetation). These models are required for orthorectification and 3D analysis.
3D Modeling, Mapping and GIS
Photogrammetry is used in GIS not just as a visual background but as a base for creating detailed maps. The unique advantages of satellite imagery like ability to cover large areas and detect temperature changes make it very useful for large scale mapping and scientific research. Roads, buildings and hydrology layers are often digitized from orthomosaics. This requires removing radiometric and geometric distortions to ensure accuracy in the output.
For example, radiometric corrections correct for lighting inconsistencies caused by sun or atmosphere, while geometric corrections correct for terrain, sensor perspective and Earth curvature.
Challenges and Considerations
Accurate photogrammetric data requires:
High quality images from calibrated sensors.
Proper overlap of images (tie points).
Good ground control for positioning.
Errors like perspective distortion or relief displacement can affect measurements if not corrected. Advanced tools and workflows simplify these corrections to get precise results.
Conclusion
As a commercial drone service company, Photogrammetry is the foundation of many of our drone services at Holland Productions. We turn raw images into usable data, bridging the gap between visualization and accurate measurement. Whether for terrain mapping, infrastructure inspections or 3D modeling, photogrammetry is an invaluable technology used on a regular basis.