The availability of LiDAR data is expanding at a rate that is out-pacing the requisite knowledge and skills needed to effectively utilize the data. Sounds like a cart-before-the-horse analogy, to coin an idiom from a bygone era.
This conundrum first came to our attention a couple of years ago when, during a roundtable discussion at a GIS forum in one of our neighboring New England states, a local government official excitedly announced that her town had just received LiDAR (or leader, to use her exact pronunciation) from the state. She went on to confide that she wasn’t entirely sure what LiDAR was but evidently that did not dampen her excitement. Remarkably, several other forum delegates jumped on the bandwagon, to use another obsolete transportation-based analogy, and shared their enthusiasm at having received data for their town while eagerly awaiting instructions from the same state agency on what to do next.
In the months that followed, it became clear that LiDAR illiteracy is not unique to small-town New England. Many GIS agencies and departments in other states, provinces, and regions throughout the world, recognizing the increased accessibility of point cloud collection technology, have proactively embarked on massive data collection projects. As a means to justify the expense of these projects, the agencies will often provide the fruits of their endeavor to eager and yet uninformed constituents and office bearers.
The aforementioned municipal officials were certainly justified in their excitement; LiDAR data is contributing to a fundamental change in how we perceive our world. Traditional mapping practices have considered the planet from an inherently unrealistic, top-down perspective. With the emergence of 3D data formats, we are now able to develop a more realistic view allowing us to interact with our data in an immersive environment and providing the impetus for the development of new cartographic and analysis techniques.
What is LiDAR?
Let’s make one thing clear, in most circumstances, LiDAR data is not a product but a raw material. It is not an end in and of itself but rather a means to an end. A commodity, if you will. Before exploring some examples of the products that can be created from LiDAR, let’s put the brakes on (yes I know, another transportation metaphor) and consider the basic structure and characteristics of LiDAR data.
Natively, LiDAR (an acronym of Light Detection And Ranging) is a vector data format, or more specifically, it is a 3D point vector format. Each LiDAR file or dataset usually contains millions, or sometimes even billions of closely spaced, randomly distributed points, with the closeness of the spacing dependent on how the data was acquired. Most publicly available LiDAR data has been collected on an airborne platform using laser transmission and receiving technology in tandem with precise position and navigation systems. Each point is attributed with an X, Y, and Z value derived from the calculated time difference between the transmission and reception of a reflected laser pulse. An aircraft flying lower and slower will create a point cloud with more closely spaced points than one flying faster at a higher altitude. Depending on how the data was collected and/or processed, additional attributes might include, a color value, reflection intensity, and the number of returns per pulse, all of which can be visualized and analyzed.
What Can You Do With LiDAR Data?
Fully utilizing LiDAR usually involves some sort of transformation process. This transformation might involve the creation of a 3D raster surface, often referred to as a Digital Elevation Model (DEM), or it might entail the automatic creation or extraction of 3D vector objects derived from the geometric patterns in an array of points. Both of these procedures will be described in more detail later. It is also possible to derive meaningful information by simply changing how the point cloud is represented. The point display can show the distribution of the different surface-type classifications; the elevation of each point above ground; variations in the density of the points; and many other characteristics.
Editing and Filtering LiDAR Data
Almost without exception, LiDAR data files will include many more points than are needed for a particular project or task. In Global Mapper, there are numerous filtering options for removing points that are outside of the geographic extent of a project area; that are considered erroneous or noise points; or that are attributed with a surface-type classification that is not required. Before embarking on any point cloud filtering procedure, it is a good idea to scrutinize the metadata for the layer. This statistical summary will provide the necessary information about the characteristics of the point cloud to allow more informed decision-making in the filtering process.
Improving the Quality of LiDAR Data
As well as removing unrequired points, Global Mapper includes several built-in procedures for recovering points that would otherwise be discarded. The most common and most powerful application of this automatic classification process is the detection and subsequent reclassification of ground points among those that are unclassified. This procedure increases the relative percentage of points that can ultimately be employed in the creation of a DEM resulting in a higher-resolution terrain model.
Other automatic classification procedures include the detection and reclassification of buildings, trees, and utility cables, which is the first step in the feature extraction process.
Creating a Digital Elevation Model
In order to perform virtually all 3D analysis procedures, a LiDAR point cloud will need to be gridded. In this context, gridding describes the process whereby the value associated with each point in an array (typically an elevation value) is used as the basis for generating a solid 3D model. This model can either represent bare earth (a Digital Terrain Model) or an above-ground surface such as a forest canopy (a Digital Surface Model). The distinction between the two is derived from the filtering and selection of the points that are used to generate the surface.
For most LiDAR users, the primary objective is the generation of a DTM, which is the platform for a wide variety of terrain analysis workflows. Without straying too far off the prescribed path (yet another transportation reference), Global Mapper offers an extensive collection of terrain analysis tools, including volume calculation; cut and fill optimization; contour generation; watershed delineation; and line of sight analysis.
The increased availability of higher density point cloud data has paved the way (OK, I’ll stop now) for a new LiDAR processing discipline. The analysis of patterns in the geometric structure of adjacent points can result in the delineation of building models, represented as three-dimensional polygons; power lines or above-ground utility cables, represented as three-dimensional lines; or tree points, derived from the collective structure of points classified as high vegetation. Global Mapper’s vector extraction tools also include a custom extraction option where 3D lines and polygons can be generated by following a series of profile views that are perpendicular to a predefined path. This tool can be used to create a precise three-dimensional model of any elongated structure, such as a curb along the edge of a street.
The impetus behind this article was to address some typical applications for LiDAR data without delving too deeply into the technical considerations or step-by-step instructions. That said, if you are sufficiently intrigued and are ready to move to the next level, you will need software in order to utilize your LiDAR data. Global Mapper has supported the import and display of LiDAR data since before the format became widely available and each subsequent release has introduced new functionality for effectively managing and processing the data.
Several years ago, Blue Marble introduced an optional module for Global Mapper to address the demand for ever more powerful LiDAR processing tools. If you are interested in the aforementioned automatic reclassification and extraction tools, you should certainly give the LiDAR Module a try.
If you are new to Global Mapper, both the base software and the LiDAR Module can be evaluated free of charge for two weeks.
Webinar Series: LiDAR Processing in Global Mapper
We are releasing a series of short webcasts exploring the use of LiDAR data in Global Mapper and the accompanying LiDAR Module. Beginning with an introduction to the structure and characteristics of LiDAR data, each video will be approximately 20-minutes in length and will cover a specific theme or topic. To receive notification of the availability of these and other Blue Marble video presentations, be sure to subscribe to our YouTube channel or follow us on Twitter.
This is the first video of the LiDAR Processing in Global Mapper series:
David McKittrick is a Senior Application Specialist at Blue Marble Geographics in Hallowell, Maine. A graduate of the University of Ulster in Northern Ireland, McKittrick has spent over 25 years in the field of GIS and mapping, focusing on the application and implementation spatial technology. McKittrick has designed and delivered hundreds of GIS training classes, seminars, and presentations and has authored dozens of articles and papers for a variety industry and trade publications.