Adventures in Mapping (Part 1)
I had the opportunity this weekend to learn and experiment with drone mapping. I've always been interested in mapping and have recently earned a certificate in Geographic Information Systems (GIS) from Penn State University. However, I had never really thought too much about creating maps with a drone.
After talking with a few people about drone mapping, I decided it was definitely something I should investigate! I started with some quick Google searching where I learned about DroneDeploy. After viewing a few of their educational webinars, I downloaded their app and got started with their 30-day free trial.
The first area that I mapped was a field near my house that is used as a motorcycle track by some of the local kids. The first step was to use the DroneDeploy app to build a mission and identify the area to be mapped. I found this to be a very simple process. Essentially, I had to draw a parallelogram and set some settings to get the desired map results. Below is a picture of the area I defined to be mapped.
The green lines represent the flight lines that the drone will fly. Essentially, the drone is going to fly along these lines with the camera pointed straight down while taking photos at regular intervals. The flight lines change as the user adjusts flight settings which include the height, the side and front "lap" (which basically sets how much overlap photos are going to have with each other), and the flight direction (angle). For my first flight, I basically used the defaults, but I recommend reviewing the educational material to learn more about these concepts. Also note that as the flight area is adjusted the flight time changes. Obviously, a larger area is going to take longer to map, so I had to keep my flight less than 20 minutes in order to be completed with one battery. The flight for this area would take a bit over 7 minutes, and considering it would take about 3 minutes for the drone to fly back and forth between its launch spot and the mapping area, I was well within the time constraints.
Now that the map was complete, it was time to launch the drone. I was a bit unclear of the process, so it took a bit of time to figure out a procedure. I ended up following these steps:
- Connect to the drone as normal using the DJI Go app. Format SD card and then do a quick flight to make sure all controls are functioning as desired. Land drone.
- Switch controller to F-Mode and open DroneDeploy on iPad.
- Follow the steps to download the flight to the drone and then launch it. Note that for the duration of the flight, I kept the DroneDeploy app opened.
The app basically takes care of everything. The drone launched, elevated to the predetermined height of 75 meters, turned, and then headed out to the designated "start location" on the map. On my iPad, I was able to see where the drone was along the flight path as well as a small live video feed from the drone's camera. Once complete with the path, the drone turned, came back to "home" and landed. Incredibly easy!!
I then reviewed the images on the SD card. I had 91 images, of which 2 appeared to be test photos taken shortly after launch, which I deleted. The remaining photos were top-down images of the mapping area.
The next step was to upload the videos to DroneDeploy.com. Again, the interface is pretty simple. Just open up the flight on your dashboard and the click and drag the photos for upload. The total upload took a few hours, after which I had to wait another hour or so before the map was complete. Essentially, this step is where the DroneDeploy magic happens. All of your photos are stitched together to create the map and the analytical portions of the map creation take place during this step as well.
The basic aerial map looks a lot like what you'd see on Google Map's satellite view, but at a higher resolution. In this image, the area with the blueish tint is the area that I mapped, while everything else is from the underlying base satellite map.
I was very impressed at how close this map lined up with the base map. In these two images, you can see how closely the images align.
An elevation map is created as well. For this particular area (shown below), the elevation doesn't vary much, but you can certainly see the tree lines and few structures within the mapped region. This is one of the areas that I was most impressed with the product, as these elevations are determined only by the uploaded photographs, the drone does not have any sensors or other apparatus to measure the distance to ground.
A "Plant Health" map is created. I'm not very familiar with how this map is analyzed or how it is put together, but it is interesting. It would appear as through the vegetation in this area isn't particularly in good shape, which isn't surprising given that it is basically an empty field in which vehicles are regularly running. I'm sure that this would be a much more useful map in agricultural applications.
Finally, one of the most interesting aspect of DroneDeploy is that it can create 3D models. This model is essentially an interactive map in which the user can zoom in, rotate, and view the map from almost any angle. Again, this map didn't have very dynamic changes in elevations, but it was still kind of neat being able to experience a 3D view.
DroneDeploy gives you the ability to share the map to clients and customers.
I'm planning on creating a second map that has will show off some of the elevation and 3D mapping in the near future, so stay tuned for a continuation of this post.
While DroneDeploy is definitely cool and I can see some great uses for it, it is not cheap. Monthly costs range from $129 to $399. For some businesses, I'm sure that it would be easy to recoup that cost, but I'm not yet to the point where I feel that DST Drones will be subscribing.
An alternative to DroneDeploy is Maps Made Easy, which I am also going to investigate. This company seems to have more of an "on-demand" pricing model, which I think would be more appealing.