Efficient and Accurate Mapping with a DJI Matrice
PPK (Post Processed Kinematic) GPS is a great way to increase the accuracy of drone work and reduce per-project costs. There are PPK units available for a number of drones including the for the DJI Matrice 100 and 600 UAV drones. This PPK unit, along with PhotoModeler’s PPK integration, provide excellent accuracy in ground surveying with minimal ground control.
To get high accuracy with a drone photo-based survey, one usually needs to place and survey several ground control points spread across the area. Ground control points are either man-made targets (e.g. large cross) that you place on the ground before the flight, or visible natural features, and both surveyed with a high accuracy instrument (total station or RTK GPS). Minimizing or removing the need for ground control points speeds up your survey and reduces cost. This can be achieved by knowing the location of the drone to high precision. This is where the PPK unit comes in.
This test was done with the Klau PPK unit – a high-quality GPS receiver mounted to your UAV drone. Typical UAV systems implement either a standard GPS unit with no corrections – provides only low accuracy position, or a real time kinematic (RTK) solution. RTK requires a stable radio link to receive base station data which is processed on the fly. RTK can be prone to GPS signal blockage and radio link outages with UAV flight dynamics, where signals are obstructed or lost over long distances.
An alternative to RTK is post processed kinematic (PPK), where all data is logged in the UAV and processed with base station data after the flight. There is no data loss or initialization loss as in RTK due to radio link limitations. All collected data is processed with similar algorithms to RTK, run forwards and backwards through the data. This ensures the most robust results possible. PPK offers greater flexibility in operations, longer range from the base, is more reliable, more accurate and easier to use.
DJI Matrice and PPK
The Klau PPK unit is available for mounting on a DJI Matrice 100 or 600. An experiment was run with this PPK unit and a DJI Matrice 100 as pictured above. The camera has 12 megapixels. The project had 89 still photos, flying height of 40m, and 6 ground control points (for use in initial calibration and for verifying measurement accuracy). The ground control points were collected on cross targets with a survey-grade RTK GPS unit.
Here are some of the photographs from the project:
The data was processed through PhotoModeler UAS software (first performing a camera calibration – needed only once per drone UAV), and then as a test with the ground control points as checks. Here is a summary table of the deltas between the solved camera positions (solved by PhotoModeler UAS) and the PPK GPS positions.
We can see the RMS error on the 89 camera positions was 8.3 cm.
Here is a summary table out of PhotoModeler UAS using a previously calibrated camera and using two of the ground control points to simulate a minimal control situation. The summary shows the deltas between the control points and the positions solved by PhotoModeler – the 4 points used as checks only for comparison, and 2 points used in processing.
We can see the RMS error on the 6 ground control positions was 6.1 cm.
Here are the 6 ground control point deltas individually. The first two in table were carried in the adjustment and the rest are check points alone.
It is great to see such high accuracy with an efficient processing method on a popular DJI drone. If you have any questions on PPK or PhotoModeler please contact PhotoModeler Technologies.