The Terrain Creator app offers an easy and intuitive photogrammetry process for turning your drone photos into an Orthomosaic and Digital Surface Model (DSM) that seamlessly loads directly into Virtual Surveyor. Terrain Creator's photogrammetry process provides a user-friendly experience to what is normally complex and time-consuming. We'll show you what options you have as a surveyor for processing your drone photos, the fastest way you can get it done, and how beneficial using Terrain Creator can be.

Terrain Creator's Photogrammetry Workflow

After you have flown your drone and collected images of your project site, you can open Terrain Creator and start the photogrammetry process. This workflow will walk you through how to:

Add Your Drone Photos

Start Terrain Creator and in the New tab, type in your project name (we recommend using the site's name along with the date). Click Create Project and the project is ready for you to add your drone photos.

  1. Click Add Photos and browse to the drone photos you want to use for your project.
    1. Option 1: Add your drone photos as JPEG files.
    2. Option 2: Add your drone photos from a .csv/.txt file.
  2. Select the photos you want to process and click Open.

Option 1: If you add your drone photos as JPEGs, they should contain geo-tags. These geo-tagged camera positions are stored in the Exif metadata of the drone photos. (This option is used as the example throughout the rest of this article).

Option 2: You can add your drone photos from a .csv/.txt file if you want to include PPK camera positions that were processed separately from the photos. The camera positions in the photos' Exif metadata (if any are included) are ignored in favor of the PPK camera positions that are included in the external .csv file. 

Once the photos are added, you will be asked to set your Project Coordinate System.

Set the Project Coordinate System

As the surveyor, you choose what Project Coordinate System you want to work in; you can set both the horizontal coordinate system and the vertical reference. The vertical reference is either ellipsoid or geoidal, but there is no need to worry about configuring a geoid grid because Terrain Creator handles that for you.

With the Set Project Coordinate System window open, you reproject the coordinates that are associated with the photos to a Project Coordinate System. In the example below, we've picked a metric measurement system for Texas South Central.

You can always change (reproject) your coordinate system before you start processing the photos. If you change the coordinate system, the camera position and the ground control points will reproject.

Review the Added Photos

You can use a laptop directly on the project site—right after you've finished flying your droneto quickly assess which photos you want to keep and which ones you want to remove. From this point, you can decide whether you need to add more photos or remove any that you don't need for your project.

There may be photos that you want to remove from your project. Reasons to remove photos may include:

  • Photos taken on the ground during drone takeoff.
  • Images may fall outside of the required project area and do not need to be processed.
  • The photos might be blurry, or the photos may be overexposed and too bright from sun reflection.

You can use both the PHOTO and VIEWER panels, along with the Show Flight path to review and remove photos. You can use the following options to remove photos:

  • Click on a photo in the PHOTOS panel and press Delete.
  • Click on a camera position in the VIEWER panel and press Delete.
  • Hold Ctrl or Shift and use left click to select multiple photos at a time on the PHOTOS panel.
    • You can also select multiple camera positions along the flight path and remove photos by pressing Delete on the highlighted camera positions.

Add Ground Control Points

A good practice is to always use GCPs for your project. To add your GCPs, click on the Add Ground Points button and import the .csv or .txt file to Terrain Creator. 

The .csv/.txt file does need to have proper column syntax in order to work. These column titles can be in any order, so long as the information stays consistent for each respective column. Your columns need to include:

  • EPSG code (optional)
  • Name (name of the GCP)
  • X
  • Y
  • Z

Typically, you have already measured your GCPs in the Project Coordinate System, however, it is a good practice to note the EPSG code in the .csv file. If you do include the EPSG code, make sure your numbers and format are set up properly. The EPSG code needs to be the first row of your .csv file and include two numbers:

  • The first number is the EPSG code for the Horizontal Coordinate System.
  • The second number is the EPSG code for the Vertical Reference System.

You only want to add GCPs at this stage of your photogrammetry workflow.

Important: If you're wondering about using Checkpoints, then those are something you'll want to do in Virtual Surveyor after you've finished creating your DSM and Orthomosaic in Terrain Creator. Adding Checkpoints at any time during the photogrammetry process will only slow you down.

Mark Ground Control Points

If you opt to mark ground control points, your initially created model improves into a corrected survey-grade model.

Note: You can process your drone photos without marking GCPs, but it is not recommended to skip this step. Read our Terrain Creator Validation article to see how using GCPs to mark and process your dataset is a quick and effective way to create survey-grade accuracy.

Prepare Your Initial Model

Click the Mark Ground Points button to prepare an initial model that helps you quickly mark your ground points.

Once Terrain Creator has created the initial model, the Ground Points panel will display next to the VIEWER, and you can begin marking points. 

The positional accuracy of your camera positions greatly influences the accuracy of the initial model. The accuracy of the model then influences how close you can get to each actual ground point position in the drone photos. 

  • Accurate model with RTK or PPK- the initial model will already be fairly precise. 
    • Terrain Creator knows where the point will be in each camera position and zooms in on the marking point in the photo.
  • Less accurate model with GPS - the initial model will be less precise.
    • Terrain Creator knows where the point will be in each camera position within a larger radius and will be zoomed out to see the marking point from a distance in the photo; you will need to zoom in more to specify the point's location.

Use Terrain Creator's Easy Guided Process to Mark GCPs

You can mark points using the guided process to improve your model.

Click on the first ground control point from the list to see the first proposed marker. Terrain Creator will automatically propose a photo as the candidate for the first GCP you can mark.

Move your cursor down to the image, zoom in to the GCP using the scroll wheel on your mouse, and mark the center of the marker with a left click. After clicking the mark in place, press Spacebar to move on to the next candidate. Repeat this process of using left-click to put a mark in place for each potential candidate and press Spacebar to go to the next photo candidate.

You may come across images where you don't want to mark the GCP from that camera position, or you need to change where you initially placed the mark. You have three options when marking points: 

  1. Simply press Spacebar without marking the photo to skip it; that way, you don't end up getting a bad measurement from marking a GCP where you cannot clearly identify the center of where the mark should be. 
  2. If you ended up placing a mark, but you don't like where it was placed, simply change where the mark is located by clicking a new location on the photo.
  3. Or, in the event you decide you don't want to use a camera position after marking it, click on the trash bin icon above the GCP image to remove the mark, then press the Spacebar to skip that camera position (as shown in the imagery below).

You Decide When to Stop Marking Points

The more GCPs you mark, the more the model improves. The more the model improves, the easier it becomes for Terrain Creator to estimate where to focus on the ground markers. You should notice the mouse curser begins to get closer to the center of focus with a closer zoom of the next candidate.

You decide when you want to stop marking points, and a good indication is when your mouse curser is landing near the center of focus for each candidate. At this point, you know your model is good and you can process your drone photos.

Generally, you should be getting better estimates as you go along. However, if the estimate for marking the GCPs is not improving and your ground points are not aligning to the center of your photos, there is no point in producing the ortho and DSM, as the results will not be good. 

Any issues should be solved before processing your photos. We have a troubleshooting guide coming soon to help you pinpoint problems you might be experiencing. In the meantime, contact us by submitting a ticket with any issues you may be experiencing.

Start the Photogrammetry Process

You are ready to process the drone photos once you know your marked GCPs are accurately matching the drone's photo locations.

Click Process Photos to start the photogrammetry process.

This process can take minutes, hours, or even up to a day depending on the size of the project, the number of photos, resolution of the photos, and how powerful your computer is at dedicated photogrammetry processing.

Results: Inspect Ground Point Accuracy

Once the processing ends, you can see the newly created Ortho and DSM in the VIEWER, combined with the accuracy information of the ground control points. At this point, you no longer need the flight path so you can turn that off and review the errors of the ground control points.

As you can see in this example dataset, the overall RMSE is already really good at .016 meters, which is just over a centimeter. You can sort the errors of the GCPs in the GROUND POINTS table on the right to check which point has the largest offset. In the example, point 11 has the largest error but even this 2.8 cm error is very low. You can double-click on the point in the list to zoom in on that point in the VIEWER.

Ground points might appear slightly under the terrain and can be difficult to see. This can easily be solved by switching to 2D mode by clicking on the 2D icon found on the lower right corner of the Status Bar.

The positional accuracy of ground point 11 is really good. You can go through and check each point to make sure they all have good positional accuracy. 

Move the Project to Virtual Surveyor and Start Your Topographic Survey

Move your project to Virtual Surveyor by clicking on Open in Virtual Surveyor

Once in Virtual Surveyor, you can load Checkpoints from your project site to verify the accuracy of your GCPs. Then begin the real work and start creating a topographic surface.

Additionally, if you already have a project open in Virtual Surveyor, you can add the newly processed ortho and DSM to your project as a Terrain State. To do this, click on the To Existing button to instantly create the new Terrain State in a Virtual Surveyor project that you have open that covers the same area of land. Switch to Virtual Surveyor to see the new Terrain State and start your work.