To test the accuracy of the Phantom 4 RTK, DroneDeploy created a control point system for a rooftop building survey. Then, 30+ flights were logged - each exploring the effect of flight altitude, overlap and number of images - and processed using DroneDeploy's Map Engine.
These data sets were then used to calculate both the horizontal and vertical accuracy of the measured points, along with the accuracy of linear measurements.
Background to the test
Over the last few years, deploying drones for construction, surveying and mining missions has become increasingly popular. And it's easy to see why more and more industries are turning to this technology.
Traditional mapping methods involved using ground-control points, which are defined points on the surface of the earth that have a known geo-referenced location.
PPK (Post Processed Kinematic) surveys are similar to RTK surveys, but the positions are not corrected in real time. It usually involves placing a stationary base station over a known control point, or a monument to allow for geolocation. GPS data is then simultaneously collected by the base station and the drone as it flies.
That data is then downloaded from the base station, and the rover (drone), and post-processed using a GPS software. These images can then be uploaded to DroneDeploy’s Map Engine for processing
When choosing between RTK or PPK methods, you need to make a choice between productivity and accuracy of the resulting imagery.
The RTK workflow can be a very quick way to obtain accurate imagery but relies on a real-time connection to produce accurate maps.
A PPK solution takes more time to set up, but relies on its signal backup data to ensure positional accuracy of the flight.
More about the drones used in the study
The primary drone model used in this study is the DJI Phantom 4 RTK which was released in October 2018. The drone was designed to use RTK processing to produce highly accurate aerial maps.
In addition to the RTK unit, the Phantom 4 RTK also uses a redundant GNSS (Global Navigation Satellite System) module that provides additional flight stability when flying in dense regions with poor RTK signal.
DJI claims 1cm RTK horizontal positioning accuracy and 1.5cm RTK vertical positioning accuracy. In terms of absolute measurements on photogrammetric models, DJI claims 5cm accuracy when flying at 100 meters with 2.7cm GSD (Ground Sampling Distance).
The RTK module, pictured below, is integrated directly with the drone to provide centimeter-level accuracy for stable flight and accurately georeferenced imagery.
A new TimeSync system was added to continually align the flight controller, camera and RTK module - ensuring each photo uses the most accurate metadata and fixes the positioning data to the centre of the CMOS sensor.
The Phantom 4 RTK also has the ability to connect to the D-RTK2, NTRIP (Network Transport of RTCM via Internet Protocol), or store the satellite observation data to be used for PPK. The D-RTK 2 Mobile Station, pictured below, was not measured in this study, instead the NTRIP method was used.
As well as the Phantom 4 RTK, DroneDeploy used the Phantom 4 Pro (pictured below) in conjunction with GCPs to serve as the control data.
Methodology for the test
The test site comprised seven different 1’ x 1’ GCPs, which were placed at varying elevations, latitudes and longitudes on the surface to be able to best capture differing scenarios.
Additionally, four checkpoints were placed on the roof and the distances between each were measured with a tape measure to establish the length between the points.
They were then tagged and measured using a Trimble RTX GPS unit to identify the lat, long and elevation of each of the points.
To capture the aerial data, DroneDeploy flew 30+ individual flights using the DJI Phantom 4 RTK and DJI GS RTK App for the RTK flights, and the Phantom 4 Pro V2.0 and DJI GS Pro app for the Phantom 4 Pro with GCP flights.
SmartNet, a third-party corrections service, was used for the RTK controller in conjunction with a mobile hotspot.
A total of 20 flights were flown with the Phantom 4 RTK at 100 feet above the surface of the roof, in both 2D and 3D photogrammetry modes and then processed with DroneDeploy’s Map Engine.
Similarly, 12 flights were flown with the Phantom 4 Pro and processed with GCPs with DroneDeploy’s Map Engine to serve as the control to compare the results against.
Slight variations in flight planning, lighting and altitude were introduced to simulate a real-world environment when collecting this data.
To capture the GCPs for the control, DroneDeploy tagged the geographic center of the targets placed on the roof with a Trimble Catalyst.
The targeted points had a horizontal standard deviation of 0.24cm and a vertical standard deviation of 0.49cm across the seven ground control points and checkpoints tagged on the test site.
These points were also used to serve as control points when measuring the accuracy of the P4 RTK against GCP measurements.
DroneDeploy would recommend using a high accuracy GNSS receiver, such as Catalyst, when processing maps with DroneDeploy to verify the accuracy of the maps on your jobsite for auditing and assurance.
Key findings
P4 RTK delivered 2cm relative vertical accuracy and 1.2cm relative horizontal accuracy
Each of the four checkpoints
This table shows the vertical accuracy of the checkpoints, both in the RTK processed map and the Phantom 4 Pro with GCPs map.
With the Phantom 4 RTK, 1.20cm relative horizontal accuracy was achieved. With the Phantom 4 Pro and GCPs, 0.90cm relative horizontal accuracy was achieved.
The study showed that the Phantom 4 RTK can achieve 2cm relative vertical accuracy, and 1.20cm relative horizontal accuracy when flown at 100 ft (33m).
It’s also important to note that the accuracy measurements for the Phantom 4 RTK were taken using DroneDeploy’s annotation and measurement tools. The GSD for the resulting maps average 0.4in/px.
Why is this important?
Determining accurate elevations is an essential part of many civil construction projects like road, bridges, and highway projects. Grade lines, drainage structures and other highway features are designed with existing, known base planes, and final elevations, or design plans.
Due to its importance in all other phases of the project development, vertical measurements are established at a control point, known as GCPs.
Due to the sizeable downstream effect of inaccurate vertical measurements, this is a critical component of pre-construction, and reducing the resulting error is of utmost importance.
Phantom 4 RTK control distances within 0.1% of Phantom 4 Pro with GCP measurements
In the Linear Measurement Accuracy of DJI Drone Platforms and Photogrammetry study, DroneDeploy demonstrated that using a Phantom 4 Pro with GCPs would produce a margin of error of about 0.64%. Using a Phantom 4 RTK, DroneDeploy was able to produce a margin of error of about 0.27% when compared to 'real-world' measurements.
As the table above shows, on average, Phantom 4 RTK measurements are within 1.09cm, or 0.10%, of Phantom 4 Pro measurements processed with GCPs.
These results show that using a Phantom 4 RTK drone, a user can expect - on average - 1.09cm accuracy on RTK processed maps in
Why is this important?
Linear measurement accuracy is an important metric to consider because of its real-world applications.
While it can be important to look at the horizontal and vertical accuracy of a single point, many civil construction projects look at grade lines, drainage along a road, volumetrics of stockpiles, or material area - all of which rely on the accuracy of point-to-point measurements.
This process involves understanding the change in accuracy when measuring the variation between points.
For example, a ±2cm horizontal or vertical accuracy could equate to ±4cm linear measurement accuracy between two points. This provides a more tangible understanding of how accuracy could affect measurements
of a road, bridge, highway or other civil construction project.
Phantom 4 RTK control distances within 0.27% of Trimble Catalyst measurements
Once the Phantom 4 RTK maps were processed within DroneDeploy, the control distances were tagged and measured across all of the maps using DroneDeploy’s measurement tools.
The comparison between the Phantom 4 RTK map measurements and the real-world measurements can be seen below.
The average error for the Phantom 4 RTK maps came out to 3.65
Looking at solely the XY accuracy of the control points, the Phantom 4 RTK delivers 1.2cm relative horizontal accuracy. These results show that using a Phantom 4 RTK drone, a user can expect, on average, 3.65cm linear measurement accuracy on RTK processed maps in DroneDeploy.
2D and 3D flights did not significantly affect measurement accuracy
A total of five
Average linear measurement error for the 2D flights was about 3.57cm or 0.26%. Average linear measurement error for the 3D flights was about 3.73cm or 0.28%. Average linear measurement error for Phantom 4 Pro flights was about 3.53cm or 0.26%.
At