8.9. Digital Photogrammetry

Modern photogrammetry is almost entirely digital, increasing efficiency and accuracy.

Digital photogrammetry refers to the modern practice of using digital technology to capture images and process them to obtain measurements and create models of physical objects or areas. This shift from traditional methods, which often relied on film and manual techniques, has dramatically enhanced the speed and precision of photogrammetric work. By leveraging digital tools, data can be processed more quickly and accurately, leading to improved results.

Imagine taking photos of a landscape with a smartphone. The digital camera instantly stores high-quality images, allowing users to edit, enhance, and analyze the photos right away. Similarly, in digital photogrammetry, these captured images provide immediate access to valuable data that can help in mapping and modeling.

Image Acquisition through digital cameras, drones, satellites.

Image acquisition in digital photogrammetry involves capturing images using various tools such as digital cameras, drones, and satellites. Each of these devices has unique advantages. Drones can access hard-to-reach locations and capture high-resolution imagery from different angles, while satellites can cover much larger areas at a broader scale. This diversity in acquisition methods allows for more comprehensive data collection, increasing the variety of applications in surveying and mapping.

Think of how different cameras are used for different occasions. A DSLR might be used for a wedding to capture high-quality portraits, while a GoPro could be used to film extreme sports activities where flexibility and portability are key. Similarly, choosing between drones, cameras, or satellites for photogrammetry depends on the specific needs of the project.

Image Matching Techniques: Feature-based: SIFT, SURF, ORB; Area-based: Normalized Cross-Correlation (NCC).

Image matching techniques are essential in digital photogrammetry to identify corresponding points and features across multiple images. Feature-based methods like SIFT (Scale-Invariant Feature Transform) and SURF (Speeded-Up Robust Features) detect and match distinct points between images, ensuring that the same features are recognized despite changes in angle or scale. Area-based methods, such as Normalized Cross-Correlation (NCC), evaluate pixel intensity patterns to find correspondences in a broader area. These techniques enhance the accuracy in constructing 3D models from images.

Consider how facial recognition works on smartphones. It identifies key facial features such as eyes, nose, and mouth (like SIFT and SURF identifying features) to verify the user’s identity. Similarly, in photogrammetry, algorithms match features across images to ensure the correct reconstruction of 3D spaces.

DSM includes all surface features (buildings, trees). DTM represents bare earth surface.

In digital photogrammetry, models are categorized into two types: Digital Surface Models (DSM) and Digital Terrain Models (DTM). A DSM captures all visible features on the surface, including buildings and vegetation, providing a comprehensive view of the environment. In contrast, a DTM focuses solely on the bare earth surface, stripping away all objects. These models are vital for various applications, such as flood modeling, urban planning, and creating topographical maps.

Imagine editing a photograph of a city skyline. The final image can either show everything in detail, including buildings and trees (like a DSM), or you could use editing software to remove all the structures, leaving just the land contours visible like a DTM. This differentiation in models serves different planning and analysis purposes.

--