Overview
Accuracy in the drone industry has become increasingly important, especially when you are intending to collect quality data for high-end projects such as photogrammetry or mapping.
However, there seems to be a lack of understanding of the different types of accuracy that drone data can provide.
In this post, we will be discussing two fundamental forms of accuracy when it comes to drones and what those different types of accuracy mean to you when you're collecting your high-quality data.
The Two Different Types of Accuracy
As we have mentioned above, there are two different types of accuracy. In this section, we will be looking at positioning accuracy and measurement accuracy.
Type 1: Geo-Referencing Accuracy (absolute accuracy)
Positioning accuracy is also known as Geo-referencing accuracy. A geo-reference is a position on the earth's surface in relevance to an exact position, in this case, your drone's position.
This is important to determine where the drone is at the point of data collection which in turn gives the data an accurate position.
Type 2: Measurement Accuracy (relative accuracy)
Measurement accuracy refers to how accurate the data is that has been collected by the drone.
This is important for post-processing the data and utilizing that data for such results as photogrammetry and stockpile measurement where the data has to correspond with measurements from a real-world environment.
So now we have looked at the two different types of accuracy let's dive a little deeper into both and look at ways of implementing them...
Type 1: Geo-Referencing Accuracy
We will now refer to positioning accuracy as geo-referencing accuracy as that is its correct definition, also known as absolute accuracy.
As we have already discussed, it's very important that the drone's positioning accuracy to its location is recorded correctly.
As standard on 99% of drones on the market, this is determined by their GPS location. This is where the drone's internal GPS (global positioning system) antenna takes its location from multiple satellites orbiting the earth.
Using these satellites it is able to accurately record its location at any time whilst moving through 3D space.
Credit: Lockheed Martin
But there are ways of improving this accuracy beyond GPS, below we will look at three different methods to increase the geo-referenced position of the drone.
GCPs
GCPs, short for Ground Control Points are points that are laid out on the ground that have known coordinates.
A GCP is often a square of chequered fabric, as shown above, that is laid down across
the site at various designated spots.
These markers are measured using survey-grade equipment, like a base station, and the positions taken are uploaded alongside the images in the processing software.
These markers or GCPs provide accurate positioning points on the ground that the drone can relate to when collecting the data.
For more information on GCPs please see our post:
RTK
RTK stands for Real-Time Kinematic and is a technique used to increase the accuracy of GNSS (global navigation satellite system) positions using a fixed base station that remotely sends out data to a moving receiver.
RTK is a real-time method of defining a drone's positioning accuracy, and it achieves this by using the drone, controller, and base station in harmony to correct and determine the drone's exact position in real-time.
For more information on RTK please see our post:
or watch our video on the subject:
PPK
PPK stands for Post Processed Kinematic and is a method of correcting the data after the flight has taken place and the data has been uploaded.
PPK works by taking the drone's data, which will be geotagged, and combining that with the base station's positional information.
Once the flight has been completed both the drone geotagged data and the base station data are matched up after uploading using your chosen PPK software. The less accurate geotagged data from the drone is corrected providing a more precise picture of your captured asset.
For more information on PPK please see our post:
So by using any one of these methods to increase your position accuracy what accuracy do you expect to achieve?
Well, it is very site-dependent but you can expect accuracy down to at least 1cm...
Now that's pretty accurate!
Type 2: Measurement Accuracy
When it comes to measurement accuracy we are referring to the accuracy of the data that has been collected by the drone. Also known as relative accuracy.
So how do we improve the measurement accuracy?
This is where GSD comes into play...
GSD or Ground Sampling Distance refers to the amount of ground/surface area covered by a single image in flight. If you're flying a mapping flight with the camera facing down (nadir position), then this distance is basically the amount of ground covered by your drone per image in flight. If you're flying vertically and mapping a tower or a facade, then GSD is basically the amount of facade surface area covered by a single image in flight.
You might see GSD frequently expressed as cm/pixel. As implied, this refers to the amount of ground or surface area covered by a drone image.
Data needs to be of a high resolution to be able to accurately map or 3D model.
Depending on what drone you are using you may need to fly closer to the target to increase the GSD, this is often the case with smaller enterprise drones such as the M2EA and the M3E.
If you are using a drone with interchangeable lenses or a camera system with a higher megapixel rate you should be able to achieve high GSD at a further distance.
Sometimes there will be a trade-off between safety and GSD, always consider safety to be your number one priority!
For more information on GSD please see our post:
So with a high GSD, you are able to utilize the measurement features in photogrammetry software as featured in Hammer Missions
Taking measurements directly from a 3D model is a very useful asset and enables you to measure stockpile volumes, cut and fill measurements, and building attributes without having to physically measure them on site.
For more information on using measurements for a 3D model please see our post below:
Or our video which discusses volumetric measurements:
So what sort of measurement accuracy would you expect?
Well, this all depends on a few factors...
The first is your GSD, but as previously mentioned, the higher the GSD the more accurate your data will be.
The second really depends on the structure or volume being measured but measurements can be in the range of x% error from the ground truth as you will have seen in our post Accuracy of Measurements in Drone 3D Models.
Can You Get Both Geo-Referencing and Measurement Accuracy?
So the question is can you get both geo-fencing accuracy and measurement accuracy at the same time?
The answer is yes! But to be able to achieve accuracy for both you will need to observe the "To Do's" for both...
So in brief, what are those "To Do's"
Geo-Referenced Accuracy:
Additional equipment is required for higher-quality accuracy:
GCP's
RTK base station, for example, the DJI RTK2
Measurement Accuracy:
High-quality data
Software, such as Hammer Missions, to be able to give accurate measurements from 3D models.
Summary
We hope this article helps you understand the difference between accuracies, furthermore, it's important to know what form of accuracy your client or stakeholder is expecting.
If in doubt, talk to your client or stakeholder to get a better understanding of their expectations and end result.
If you'd like to learn more about how to produce high-quality data and get the most out of your drone missions, please feel free to visit our learning resources
If you haven't got a Hammer account as of yet and would like to try Hammer Missions you can get started on our free trial.
To learn more about our enterprise solutions, including mission collaboration, data processing, and AI solutions, please contact us at team@hammermissions.com
We look forward to hearing from you.
- Team at Hammer Missions