Breaking down drone mapping: step-by-step guide with Tim Durham

Drone photogrammetry is used in various industries, from land surveying and construction to archaeology, real estate, insurance, and forensics. As technology advances, aerial mapping is becoming increasingly affordable and accessible. Innovations in drone technology and specialized software are making the surveying process faster, easier, and more efficient.

Despite its accessibility, the technology remains a complex field with many nuances. In this blog post, we’ll walk you through the entire RTK drone mapping workflow from scratch, using an example from Tim Durham, the owner and chief pilot of Midsouth Drone Services and Drone Mapping Tools.

Tim demonstrates how to create a 3D model of an old barn scheduled for demolition, preserving its structure digitally. For this, he uses the Emlid RS2+ base, Reach RX rover, and a DJI Matrice 300 RTK drone equipped with a P1 camera. While 3D mapping involves additional photos, the basic principles are the same as standard drone mapping.

What is drone mapping

Drone mapping is a technique that uses drones with cameras to survey areas and gather highly accurate geospatial data. To begin, you plan the flight mission in specialized software to ensure the drone covers the entire area. During the flight, the drone’s camera captures images of each feature at set intervals. The goal is to create an orthomosaic map—a seamless, detailed map stitched together from these images in post-processing software.

To ensure precise alignment, you also need a minimum of 5 ground control points (GCPs) and more for larger areas. GCPs are specific, marked locations on the ground with known coordinates, determined through RTK surveying either before or after the flight. These targets should be designed to contrast with their surroundings and be large enough to be clearly visible in the drone’s images. They generally are no smaller than 75 cm (18 in). Acting as reference points, GCPs help accurately align and position the map within the real-world environment.

Requirements for setting up the drone mission

A basic setup for a drone mapping mission usually includes:

• A base station to send corrections to the RTK drone,
• An RTK drone for aerial mapping,
• A rover unit for collecting GCPs,
• 4-5 ground control points (GCPs) to accurately align and position the map within the environment. Larger areas require more.
• Drone mission planning and flight control software.
• Photogrammetry software for processing digital images and generating 3D spatial data.

However, every project has its unique requirements, demanding different setups. Tim’s setup includes an Emlid RS2+ base to send corrections to his DJI Matrice 300 RTK drone with a P1 camera, a Reach RX rover to capture GCPs, several GCPs, DJI flight planning software, DJI Terra photogrammetry software. You can learn more about Reach integration with DJI RTK drones in our documentation.

Drone mapping in action

Now, let’s go over the entire process from scratch using Tim’s example of preserving the old barn’s structure in the digital format before demolition. While some steps may vary slightly, the core process remains the same for all aerial mapping activities.

Setting up the base

Upon arriving on-site, your first task is to set up the base that will stream corrections to the drone. For precise georeferenced data, which is linked to a specific location on Earth using a coordinate system, you need to position it over a known point. This may be done using an already established control point or, in most cases, setting it up yourself. To learn more about the base setup methods, check out the guides in our online documentation.

Tim sets up his Reach RS2+ base as a rover to receive corrections from a Mississippi CORS network (NTRIP) and then collects the point using the averaging option for 8 minutes. This control point provides fixed ground coordinates, linking aerial images to real-world locations. This step ensures the map aligns accurately with other geospatial data, ready for integration into mapping and GIS systems.

Preparing GCPs

While collecting the control point, Tim sets up several GCPs around the barn area. These targets, contrasting with the surrounding area, will stand out in the drone’s images and help accurately align and position the map within the real-world environment. For targets placed around vehicles or human activity, Tim paints small marks on the ground by the GCP corner to verify they have not been moved since their placement.

To achieve high accuracy and seamless map alignment, you should keep the following in mind when placing GCPs:

• Use a bi-pod for a survey pole: Even the slightest touch of the survey pole can cause the rover to move while recording the point’s coordinates. It’s better to use a bi-pod to ensure the rover stays still while recording the point. Alternatively, you can use Reach RS3 with the tilt compensation feature and let it eliminate such errors.
• Take 5–10 seconds when collecting the point: When you record the coordinates of your GCPs, average their positions for 5–10 seconds. Reach receivers provide a reliable fix solution, but it’s better to ensure that the FIX solution is updated accurately.
• Use enough GCPs: Depending on the survey area size, plan for about 5 GCPs.
• Ensure visibility: GCPs should be large and highly visible, contrasting with their surroundings. Each GCP’s center should also be clearly marked.
• Distribute GCPs evenly: Space GCPs across the area. For terrain with elevation changes of 50–100 meters, place GCPs closer together to account for variations. If using 5 points, place one at each corner and one in the center.
• Create your own checkpoints on-site: Besides the GCPs, it is also useful to check your data with some permanent distinguishing features of the site. It can be an easily seen corner of the pavement, or something else. If there are no “natural” distinguishing features on the site, create your own. For example, it could be a rebar that’s been pounded into the ground to act as a control point for check-in and check-out shots. Painting the GCP or сheckpoint on the ground or pavement is an efficient way to create a target without having to return later and pick it up.

Сollecting GCPs positions

The next step is collecting GCPs positions—this can be done either before or after the flight. The key is to gather them in the correct coordinate system and with a FIX solution. They will be needed for data processing in the photogrammetry software later. Tim completes this step post-flight using the Reach RX receiving corrections from the Mississippi corrections network. You can also do the same with Reach RS2+. Check out the Preparing ground control points for PPK UAV mapping guide to learn more about GCPs placement and measurement.

Setting up an RTK connection between the base and the drone

To capture imagery with centimeter-level precision, you need to establish an RTK connection between the base and the drone. Tim uses a manually defined point as his base’s position, which he previously collected using the averaging feature, and emphasizes the importance of double-checking it before the flight. After that, he sets up the correction link between the base and rover via Emlid Caster for accurate positioning.

When using an RTK connection through Emlid Caster, make sure there’s reliable cell service in the area. If cell service isn’t available, consider using a local NTRIP option as an alternative.

Both ways are described in the DJI RTK drone and Reach RS2/RS2+ base integration guide.

Setting up drone flight

The next step is configuring your drone’s settings in drone mission and control software. Tim does the following:

1. Double-checks that his DJI drone is receiving corrections from the Reach RS2+ through Emlid Caster.
2. Sets up a mapping mission by defining the area he wants to cover with an added margin to ensure the ground control points are included.
3. Selects his drone’s camera model.
4. Selects the oblique collection mode for 3D mapping. In this mode, images are captured at an angle instead of directly downward. By contrast, a NADIR image is taken with the camera pointing straight down below the drone.
5. Sets the flight altitude and angle course.
6. Sets the photo overlaps: front overlap to 80% and the side overlap to 75%. The front and side overlaps determine how much each photo overlaps with the next during post-processing. Increasing the side overlap improves stitching precision by capturing more matching features but reduces the area covered in a single flight. Increasing the front overlap speeds up photo capture but may eventually reach the camera’s technical limits.

All these settings are adjusted based on the specific conditions of the flight mission.

Now, everything is ready for the flight. Tim performs two flights—one following the configured mission and another manually, descending from the set altitude to a lower level to have a good representation of the barn facade features and details.

Processing the data

As a result of his fieldwork, Tim has the following data to process back in the office, which will later become a 3D model of the barn and the surrounding area:

• Set of images with EXIF data from his DJI drone
• CSV file with the GCP positions

One of the most important things to keep in mind when post-processing the resulting data is setting up the correct coordinate system for the project. The choice should be based on the base’s coordinate system and considered both when importing images and GCPs positions.

Tim imports the images with EXIF data from his drone into DJI Terra for photogrammetric processing. Then, to improve the project’s overall accuracy, he exports the EXIF data and edits the horizontal and vertical accuracy values to 1 meter in Excel, ensuring the software prioritizes control point accuracy over image data. While this adjustment isn’t standard, it’s a common practice in photogrammetry to ensure that control points guide the processing, resulting in better overall accuracy.

Optimizing precision with GCP data

Afterward, he uploads the corrected EXIF data and imports the GCP data collected with Reach RX, which will take precedence over the EXIF data to further enhance the project’s accuracy. When working with GCPs, Tim ensures their positions are recorded accurately and, if necessary, optimizes them using DJI Terra’s tools. The GCPs are then matched with the data in the images and used to make fine adjustments, aligning the image coordinates with the ground control points. Additionally, Tim uses a checkpoint, which is easily seen as a fixed object in the surveyed area, to visually monitor for any distortion when getting the ready model.

To ensure that the whole area is covered, Tim also uploads a KML file with the surveyed region of interest to filter out any excess data, which he created while collecting GCPs with Reach RX.

Finally, the 3D model is rendered with detailed, immersive views, which can be shared online via DJI Modify.

Video tutorial by Tim Durham

If you prefer video, check out Tim’s tutorial, where he breaks down the entire 3D mapping process step-by-step.

Which Reach receivers to use for drone mapping

The Reach RS2+ and Reach RS3 base stations are perfect for an RTK drone. They offer two integration options. Reach can send NTRIP corrections via Emlid Caster through the internet or using the Local NTRIP option in Emlid Flow without connection to the internet. You can also log data for backup in case some images are taken in the Float or even a Single solution due to electromagnetic interference during flight and post-process it in Emlid Studio to improve the solution.

To collect GCPs, you can use Reach RX or Reach RS3. Reach RX is a perfect choice if you work in an urban area and need something light and compact. It accesses network corrections using the internet connection on your smartphone or tablet so that you can collect ground control points within an instant. With Reach RS3, you can work in any area. It receives corrections from an NTRIP service via LoRa radio and from third-party bases supporting the TrimTalk protocol over the UHF radio. Reach RS3 also offers the tilt compensation feature that simplifies stakeout and provides survey-grade results even in hard-to-reach spots.

Order your Reach from the Emlid online store

Experience the power of Emlid Reach receivers for aerial mapping! Order yours today from the Emlid Store—now available for immediate shipping.