The Lidargrammetry section of Chapter 7 of Maune (2007) describes the creation of pseudo stereo pairs (PSP) from lidar elevation and intensity data. These images can be set up for 3-dimensional viewing in a digital photogrammetric workstation (DPW). The CAD and GIS drawing tools that are integrated into the DPW can be used to extract 3-dimensional features such as buildings, road edges and centerlines, ridge and drainage features in the terrain, and breaklines for hydrologic enforcement. In stereo photogrammetric collection, the human operator visually interprets elevation by clearing x-parallax; the ability to do this accurately and consistently varies from person to person. In lidargrammetry, the elevation is derived from the lidar data itself, thereby eliminating uncertainty inherent in human interpretation. Conversely, human operators can make manual edits to lidar point classification at the same time they are performing feature collection, saving time and streamlining the production workflow. Finally, 3-dimensional features collected by lidargrammetry will match other lidar-derived products, ensuring consistency in all of the final mapping and GIS deliverables.
In traditional photogrammetry, digital terrain models are constructed of individual mass points (equivalent to the individual lidar points) and breaklines, which define sharp edges and elevation discontinuities in the natural or man-made landscape. Including breaklines when generating a TIN, for example, ensures that the resulting surface depicts these edges in a realistic way. While lidar is great for collecting millions of individual-accurate elevation points, it is not inherently capable of "connecting the dots" to establish the topology of linear features. A roadway, bridge abutment, or retaining wall may be detectable in the lidar point dataset, but the edges of that object will not be clearly delineated as if they had been manually digitized as a three-dimensional line.
In the early days of topographic mapping with lidar, data providers struggled to find ways to add breaklines to lidar point datasets. One approach was to use conventional photogrammetry to extract breaklines from stereo imagery, and then combine these with the lidar points before generating TIN's and contours. There are several rather obvious drawbacks to this approach. First of all, it requires acquisition of an additional dataset with a second sensor that has different operating requirements and constraints. This adds cost and time to the project, and reduces many of the advantages of using lidar in the first place. Second, it requires precise registration of two independent datasets at accuracies equal or better to the desired end-product. Accomplishing this requires integration of two otherwise independent data processing workflows, which again translates to added cost, time, and training for production technicians.
A geospatial software development company, GeoCue, developed a technique for displaying stereo pairs of lidar intensity images using a standard digital photogrammetric workstation. This allows mapping technicians to view the lidar just as if it were stereo imagery, so they can manually digitize 3D lines and polygons. This important technological advancement makes it possible to use lidar source data simply and effectively in a traditional photogrammetric workflow for the extraction of supplemental breaklines and feature extraction. Little additional training is required; an accomplished stereo compiler can become a skilled "lidargrammetrist" in a matter of a few hours. We will not be able to demonstrate lidargrammetry in this course, because it requires special hardware for stereo display. However, you will use lidargrammetrically-derived breaklines in the lab activity to see how they affect generation of contours and other topographic analyses.
You will read more about lidargrammetry in your textbook reading assignment. You can also explore these additional articles: