Global Mapper Pro can thin point clouds, classify, and extract features. Before you begin, be sure to make the necessary adjustments to the point cloud’s location as measured by positional accuracy. Erroneous offsets in or between point clouds can impact analysis and surface generation. In this context, positional accuracy refers to how close the measurements are to real-world values. Confidence in positional accuracy is a vital component of point clouds as the clouds are often the basis for analysis and derived data products. Adjusting a point cloud to match the position of ground control points (GCPs) collected with a high-accuracy GPS unit, or fitting it against another overlapping point cloud with an established position, can help mitigate positional accuracy issues. These tools, alongside many others in Global Mapper Pro, can be used seamlessly with aerial and terrestrial lidar or photogrammetrically derived point cloud data.

Data Used in this Scenario

Here we have a terrestrial lidar data set captured at ground level along the main street of Hallowell, Maine. As you can see in the image below, this ground-level collection method was unable to capture the tops and rear sides of buildings. To fill these gaps in the data, the point cloud is going to be merged with an aerial point cloud from the USGS. This aerial cloud has a lower resolution but will still provide context, and elevations. Additionally, USGS point clouds go through a thorough quality assurance process before publication, and as such, they are known for their high positional accuracy.

This scenario will showcase two methods of improving point cloud positional accuracy: one by comparing against a USGS point cloud and another against collected ground control points. 

Collecting Ground Control Points  in the Field with Global Mapper Mobile Pro

Aside from measuring accuracy by comparing one point cloud against another, positional accuracy can also be derived by comparing a point cloud against control points collected in the field. GCPs, or Ground Control Points, are point features with X, Y, and Z attributes used to geographically reference other data. Sometimes they’re used to add spatial reference information to unreferenced data, but in this case, they’re used to improve the spatial accuracy of an existing referenced data set. 

In order to improve the accuracy of the data, the GCPs will need to be collected with a unit that is more accurate than the data. For example, the built-in locational device on most phones is only accurate to about 3 meters and would not improve a high-accuracy dataset. In this use case, the Emlid Reach RX with RTK was paired with Global Mapper Mobile to measure GCPs throughout the dataset. For more information on this setup, read this blog on High Accuracy Data Collection with Global Mapper Mobile Pro and Emlid. Global Mapper Mobile is compatible with a wide variety of different GPS and RTK units; for more information, consult the full lists for Android and iOS. 

Points Should be Clearly Visible in the Point Cloud

Horizontal adjustments require manual positioning. When in the field, be sure to place the GCPs on features that you will be able to precisely pinpoint in the lidar data, such as the corners of curbs or buildings. Lamp posts are an example of things that don’t make great horizontal GCPs, as you’ll have to remember exactly which side of the pole you measured from, as there is no clear center at the centimeter resolution. This is true even when referencing other data types, such as imagery. Using Global Mapper Mobile, we were able to take our data into the field to assess data bounds and double-check that the features being mapped were visible in both datasets. The importance of point visibility is demonstrated during horizontal rectification, as outlined later in this scenario.

Aligning Point Clouds in Global Mapper Pro

First, be sure to classify the point cloud(s) so that similar point types can be compared. We don’t want to measure grass against ground points! Once classified, the Filter Lidar tool can be used to turn off classes that aren’t to be used. Generally, positional adjustments are based on ground points. Non-ground points that have been “turned off” will be shifted with the rest of the point cloud, but they will not be taken into account during processing. 

Measuring the Distance Between Two Point Clouds

Knowing the offset distance between two point clouds or the point cloud and the GCPs isn’t required, but it does highlight any possible problem areas, which can provide clarity on how they can be adjusted. Use the Compare Point Clouds tool to measure the offset between the point clouds. Offset distance for each point is assigned as an attribute to allow the visualization of offsets across the datasets. Because Global Mapper doesn’t alter the original point cloud, you also have the option to reload the original file from your computer to measure against a shifted layer after processing. For more information on the Compare Point Clouds tool, check out this blog detailing the updated tool in version 24.1.   

Automatically Align Two Overlapping Point Clouds

The Fit Point Clouds tool automatically aligns one point cloud with another, minimizing the X, Y, and Z differences between the point clouds. Unlike the Shift option, which shifts an entire layer by the same amount, the Fit Point Clouds tool adjusts individual point features. It’s easy to use, but provides options if you prefer to stick your hands in the settings to control how the points should be adjusted. To get technical, by using an iterative Closest Point method, this tool calculates a 3D affine transformation, finding a 4X4 matrix that best fits the data. After the transform has been applied to the points, the process is repeated. The result is a best-fit transform that is as close as the ICP process gets based on specified user settings. 

In our example dataset, the Fit Point Clouds tool was able to bring the terrestrial data down to our control USGS dataset. Another possible use is as a manual way to align flight lines, as long as the point clouds are saved in separate layers.

Aligning a point cloud against GCPs

The difference in resolution between the two datasets in this example could cause issues when aligning them against one another, so GCPs may be a more accurate method. The Ground Control points were collected with a high-accuracy GPS unit, and Global Mapper can adjust the point cloud to fit the GCPs with two tools: One for horizontal alignment and another for vertical. 

Horizontal alignment can be a bit more complicated than vertical when aligning to GCPs. Global Mapper requires some user input to specify matching features between the point clouds. The Fit Point Clouds tool also has a manual option that gives you the ability to specify exactly how the point cloud aligns with the control points. This method can also be applied in 3D instead of horizontally, but I personally prefer to separate the functions to limit where error is introduced. In the image below, the GCP on the right is mapped to the point cloud in the center. With this method, there are no iterations, and the final alignment can be readjusted easily by reopening the tool. 

Vertical alignment is (relatively) easy; there are only two directions, up and down. Global Mapper’s Lidar QC tool handles this with ease. It has two functions: comparing the point cloud against the GCPs to estimate distance and error and performing a vertical rectification process on the entire point cloud based on the ground control points. The vertical adjustment is a comparison of elevation values from ground control points against the elevation values of the lidar points in the immediate vicinity of each control point.

Once the points are aligned, they can be used for further processing, such as creating surface models, classification, and feature extraction. For further information on tools for quality control of point cloud data, check out the Global Mapper Knowledge Base.

WORK MADE EASY WITH GLOBAL MAPPER

Global Mapper provides an innovative way for professionals involved in agriculture and other industries to perform a terrain suitability analysis for a variety of use cases. A few freely available data layers were used to identify areas suitable for vineyard development. Of course, not all site selection criteria can be analyzed in a GIS program. Site visits, advanced soil sampling, planning, and infrastructure implementation are all needed before beginning grape cultivation. The areas identified by this suitability analysis are now vector features with attributes describing slope, aspect, area, and soil type that can be further edited, exported, or taken into the field for further site exploration.

Want to try Global Mapper? Sign up for a 14-day free trial. You can also request a demo from one of our experts to see this workflow or other Global Mapper processing abilities.

Learn More