Volumetric Analysis with EVO
Author: Lewis Graham
Published On: March 04, 2020
Raw Data to Colorized Point Clouds
True View Evo, based on our widely-used GeoCue LP360 software, is bundled with every True View 410 3D Imaging Sensor (3DIS), including our subscription program. True View Evo contains all of the workflow tools (including embedded Applanix POSPac drivers) for taking True View 3DIS data from the USB data stick containing raw observation data to a gorgeous, geocoded 3D colorized point cloud. However, this is just the beginning for many projects. While the colorized point cloud is a fantastic visualization tool, most projects aim to produce derived products. No worries - True View Evo contains a plethora of tools for creating a wide variety of products from point cloud data. In this issue, we will take a look at basic volumetric analysis.
Volumetric Analysis in True View Evo
Volumetric analysis within a point cloud setting is the process of using the point cloud to compute the volume of stockpiled material. True View Evo contains an advanced set of volumetric analysis tools that allow for a variety of computational scenarios such as:
- Simple stockpile with base defined by a "toe"
- Simple stockpile with overhead features (that require removal from the computation)
- Stockpiles with a priori base level (for example, a surface is defined by a survey performed before the stockpiles were placed)
- Volumes with bases defined by geometric shapes (e.g. material contained in bins)
- Borrow pit analysis
- Change analysis over time such as cut and fill computations
Analyzing Volume of Stockpile Scenario
In the example of this issue of GeoCue's True View Bulletin, we will take a look at a simple situation where we want to measure the volume of a stockpile that has an overhanging conveyor. This scene, collected with a True View 410 3DIS, is depicted in Figure 1. Note the useful display modes in True View Evo that make the situation with the stockpile easy to visualize. I have set the display to render by Triangulated Irregular Network (TIN) and shade by elevation color bands. Notice how we can very clearly see the stockpile as well as the overhead conveyor.
Figure 1 - Stockpile with overhanging conveyor
The process of computing the volume of this pile will follow a few simple steps:
- Create a stockpile "toe"
- Change the classification of the conveyor points so they will not be included in the stockpile volume calculation
- Classify any extraneous "noise" points - this is often required if the point cloud was photogrammetrically derived. It is seldom needed with True View 410 LIDAR data
- Compute the volume
True View Evo has a robust set of tools for manually and automatically digitizing stockpile toes. In fact, the automatic stockpile toe creation tool includes an option to detect and classify non-stockpile overhead features. If the overhead structure is separated from the stockpile, the classification is "clean" and you will not have to do any additional work. If the overhead structure touches the pile (e.g. a conveyor at pile level), you may have to do some manual cleanup. This is very straightforward using the variety of manual point cloud classification tools within our processing software, True View Evo.
Creating a Stockpile "Toe"
In the first step of our example, I will use True View Evo's Automatic Stockpile Toe digitizing tool with an option set to move overhead points into a class called "Conveyor." The only interaction needed for this tool is to select the tool icon and click on the stockpile you wish to process. Note there are some user-definable tuning parameters, but once set for a particular data type such as the True View 410, these usually do not require per-project changes. The stockpile of our example is depicted in Figure 2. The top portion of the figure shows a Map View (plan view) of the pile and the lower section shows a profile view. Note the conveyor overhanging the pile. This is an example where the overhead structure (in this case, a conveyor) is cleanly separated from the stockpile. As an interesting side note, notice how cleanly the True View 410 laser scanner depicts the material falling from the conveyor to the stockpile!
Figure 2 - Stockpile in plan view (top) and profile view (bottom)
Toe Extraction and Conveyor Classification
To perform the toe extraction and conveyor classification, I selected a tool in True View Evo called the "Toe Extractor" with an option checked to automatically classify overhead points. I then simply clicked on the pile I wanted to define. The result is shown in Figure 3. We have a special rendering mode in True View Evo that allows points of one set of classification to be rendered as a solid model (TIN) and others to be rendered as Points. The upper right panel of Figure 3 shows this result in 3D. I have rendered all classifications except the conveyor points as a solid TIN. The conveyor points, on the other hand, are routed to the Point display. This provides a very nice inspection view for analyzing the result of the automatic toe definition and classification algorithm. You can see that within the toe boundary, none of the solid TIN model of the surface extrudes up to the conveyor. Examining the views of Figure 3, it is clear that:
- The toe is correctly placed, forming a clear and accurate circumscription of the stockpile
- The overhead material (a conveyor in this example) has been cleanly separated from the stockpile material by classifying its points to the conveyor class.
Figure 3 - The automatically defined toe with automatically classified conveyor points
When True View Evo creates the toe, it uses the underlying point cloud to extract the elevation value of each vertex. This means the toe is correctly rendered in 3D, directly on the surface. When this toe is subsequently used in a volume computation, the 3D vertices are used to compute a triangulated surface. This approach ensures the correct elevation is used at all locations when computing the stockpile base.
This toe definition and overhead point classification required about 7 seconds to run on my laptop computer. If I had to do this task using manual tools, it would take about 20 minutes.
Compute the Volume
The final task is to compute the volume of the stockpile. This is a separate tool in True View Evo that supports a wide variety of volumetric scenarios. The interface for the volumetric analysis tool is depicted in Figure 4. It is cleanly laid out, allowing you to easily define the various options for computing volumes. In our example, we choose the toe polygon as the volume "base" and the point cloud as the "hull." The volume tool creates small prismatic volume sections between the base definition and hull and sums up these elements to compute the overall volume.
True View Evo Volumetric Analysis Results
The result of running the volume computation task is shown in Figure 5. I had the Volumetric task render a "cut-fill" image while performing the volumetric analysis. This image is rendered green everywhere the point cloud is above the toe and red anywhere the point cloud dips below the toe. I have indicated a small depressed area using a red arrow in Figure 5. This cut-fill image is an extremely useful QC tool. For most stockpile computations, you would not expect to see large areas of depression. As you can see in Figure 5, only a few small areas are below the surface defined by the toe. The 3D rendering and the cut-fill image give me a high level of confidence in my model definition. The volumetric computation tool creates a polygon from the toe definition and attributes this feature with the volume results. This is also shown in Figure 5. Note the volume of this particular example was computed as 84.28 m3 (settings on the volume task allow you to choose between metric and imperial units).
Figure 5 - The computed volume
True View Quality
True View Evo has a rich collection of tools aimed at three important goals:
- Accurate results
- Ease of use
- Fast execution
I trust you can see from this example that we have certainly achieved these goals for our volumetric analysis tools. For a typical 10 acre stockpile area of 50 or so piles, you can save 8 hours of production time as compared to manual definition methods. Our overall aim with the True View Ecosystem is to put quality tools in your hands that significantly improve your overall return on investment.