Capture and process RTK data with the M300/M350

New
  • Published on May 12, 2026
  • 3 minute(s) read
  • JS
  • Matthew Toohey
 Prev Next

This article explains how to capture RTK (Real-Time Kinematic) data using a DJI M300 or M350 drone with Hovermap, and process it in Emesent Aura to produce a georeferenced point cloud.

What you will need

  • DJI M300 or DJI M350

  • Hovermap (ST, or ST-X)

  • GNSS mobile receiver

  • RTK base station: Emlid RS2+ (recommended) or DJI D-RTK2

  • (Optional)4G SIM Card Dongle (DJI Matrice 300 PT17, ZTE MF833V) for M300 Smart Controller, where a CORS network is available

If you are using Mapping mode (AL0), the onboard DJI SDK must be configured on the M300/M350 before proceeding.

Procedure

Step 1: Set up your base station

Select one of the following base station options based on your available hardware and site conditions.

Option

Method

Key considerations

Emlid RS2+ (local)

Position over a surveyed point, or use RTK Fix via CORS network

Controller must remain within approximately 20 m of the RS2+. Emesent Commander must run on an external tablet for M350.

DJI D-RTK2 (local)

Position over a surveyed point

No 4G or Wi-Fi required. Emesent Commander can run directly on the M350 RC Plus controller. Requires a surveyed point for accurate absolute position.

4G Dongle (M300 only)

Direct connection to CORS network

Requires 4G connectivity. Accuracy decreases by approximately 1 mm per km from the nearest reference station.

M350 RC Plus direct to CORS

Direct connection to CORS network via Wi-Fi

Wi-Fi at the mission site. Emesent Commander must run on an external tablet. Accuracy decreases by approximately 1 mm per km from the nearest reference station.

  1. Set up your chosen base station at the mission site.

  2. Confirm the base station is transmitting corrections before proceeding.

Step 2: Plan and fly your mission

RTK requires a clear GNSS signal. Do not use RTK for underground or indoor environments.

  1. Plan your flight path using a parallel track pattern, where the drone flies adjacent rows in one direction, makes a 180-degree turn, then flies back along the next parallel row. Set traverses no more than 30 m apart.

  2. Set your flight speed and altitude according to your Hovermap model:

Hovermap model

Recommended speed

Recommended altitude

ST-X

6 m/s

30 m

ST

3 m/s

20 m

  1. If RTK signal is poor during the mission, add more zigzags and overlaps to the flight path to improve point cloud accuracy.

  2. Fly the mission. Hovermap records RTK data alongside LiDAR data during the flight.

If the mission area exceeds the drone's battery life, some customers have reported success swapping batteries mid-scan. Emesent has not validated this practice and cannot guarantee results for all customers. If battery swaps do not produce a usable result, use third-party software to merge multiple RTK datasets.

Step 3: Process RTK data in Emesent Aura

  1. Open Emesent Aura and load your bag file.

  2. Emesent Aura detects RTK data in the bag file and prompts you to use it for georeferencing.

  3. Toggle on Use RTK data. The Processing Settings panel opens.

  4. Review and adjust RTK and output parameters as required.

  5. To change how the point cloud is georeferenced, update the Georeferencing Mode setting.

  6. Click Save to confirm your settings. To exit without saving, click Close.

  7. Run processing. The georeferenced output is saved to the write_global_landmark_referenced folder.

Advanced feature matching is disabled by default for RTK data processing. This reduces error distribution and produces a more accurate output for RTK data.

Outcome

After completing these steps, Emesent Aura produces a georeferenced point cloud in the write_global_landmark_referenced output folder. The output includes all standard SLAM output files and a projection file (.PRJ) that defines the coordinate system and projection information.

Additional information

  • For answers to common questions about RTK technology and the processing workflow, refer to Emesent Aura M300/M350 RTK FAQ.

  • RTK uses a Transverse Mercator projection with unity (1.0) scaling and a local origin calculated from RTK values. Point cloud coordinates will have smaller values than UTM and no scaling is applied.

  • Ellipsoid height is used for vertical positioning. Apply a GEOID correction using third-party software if required.

  • If your RTK system uses a datum other than WGS84, edit the .PRJ file to update the datum string.

  • Parts of the flight trajectory that lack reliable RTK data are excluded from RTK-based processing. Those sections fall back to SLAM only.

  • It is not currently possible to use both RTK and GCP corrections on the same dataset.