Drone Based GCPs
Could a small drone with an aerial target on its back be used as mobile ground control for aerial survey work?
In this series we're going to find out.
We break the task into three main problems:
What logistical issues are solved/created by this method versus traditional methods?
Can the drone's on-board GPS be used to record the GCP locations? To what accuracy?
Is this cheaper or better than traditional methods?
Traditional ground control is done by licensed surveyors with expensive GPS units. A target large enough to be visible from a manned aircraft is placed so that it can show up in the imagery (image: right). These GPS units have typical accuracy in the single centimeter range and have to be placed throughout the survey area manually. As can be seen in the image there is quite a bit of work to set up 10 or 15 of these stations. With a price from $6,000 to $10,000 per unit the motivation to reduce this complexity is clear.
Unmanned aircraft can fly much lower than traditional manned aircraft aerial survey altitudes. So the ground targets can be much smaller. Flying a second drone to drop targets and record locations ahead of the photography drone's flight has the potential to dramatically decrease target deployment time and cost. A single drone that could act as a GCP deployment platform (or act as a GCP itself) would be less than $1000. So that's about an order of magnitude decrease in hardware alone if the onboard GPS can be used for GCP location. And the drone can fly to the GCP point much faster than a person can walk the larger traditional equipment. So there's a potential to reduce labor costs for many applications.
A major downside is terrain dependencies. If there's a bush or a uneven spot as the drone lands, it could tip over. Now the operator has to walk out and retrieve the drone since it can't take off again which could negate some of the advantage. And managing multiple aircraft on one project is something that even a skilled operator would have to practice a few times. Still, there is enough potential cost and time benefit to make it worth the effort of testing.
So that's exactly what we did. The next post in the series will cover the answer to bullet point 2: Can the drone's on-board GPS be used record the GCP location and to what accuracy?
Our results were surprisingly encouraging. Math and a full write up to come in Part 2