What the heck is drone data anyway?
Drone Data – What Is It?
This article introduces the backbone concepts of this whole website. Drone data is reinventing how we map, visualize and analyze the world around us. Gathered with a drone that costs as little as $300, it opens the world of remote sensing to everyone. This data includes high-resolution aerial imagery and videos. It offers expansive and detailed perspectives of landscapes, architectural structures, and diverse terrains. But its value extends beyond visual images. It also integrates vital information such as precise GPS coordinates, and altitude to enhance the geospatial context of the images.
This enhanced nature of data collected by drones makes it versatile and meaningful across many sectors, including agriculture, construction, environmental monitoring, and urban planning. It isi the newest frontier of geospatial analysis. It dramatically improves the precision, detail, and scope of what can be analyzed spatially. This data is not just transforming; it’s fundamentally redefining the ways professionals and industries collect geospatial data for their particular application. But don’t worry, it’s very accessible to anyone who’s interested. The whole goal of this website is to make it accessible to anyone.
Let’s get down to the nitty gritty though. Data from drones comes in multiple flavors that you need to know about. Those flavors are raw data such as JPGs, TIFFs, Videos and LIDAR point clouds. Since this is an introduction to the subject and LIDAR is advanced and very expensive, I’m going to skip over LIDAR for now.
What is the “raw” drone data?
The raw data is actually the only thing that you use your drone to collect. The rest of the process is using various software products to enhance and analyze. To make it simple, this is the aviation part of the whole process. As I mentioned earlier, raw data comes in various formats including JPG, TIFF and Video formats. You collect them using your drone’s sensor, and the sensor you have determines which formats you will be collecting. However JPGs and TIFFs are both just a bunch of images your sensor takes. Think of it as taking a bunch of pictures with your camera that overlap. Videos of course are videos of what you see while flying.
Still Images (JPG, TIFF)
Still images are what you expect, pictures you take as you are flying. However when you’re mapping you have to collect them in a very special way. Another article will explain in depth how you need to collect them and I will update this to put to it. However the gist of it is that you need to point your camera downward and overlap the pictures both on the end and the sides. Think of it as taking a bunch of overlapping pictures from any kind of camera you’ve ever seen.
There is a lot of variables to collecting these images and as mentioned earlier I will go in depth in another article which will be linked here. Camera focus, white balance, shutter speed and landscape lighting all play a role in collecting good data. However once you collect all these overlapping pictures you can run them through photogrammetry software (another article will go in depth on these) and create all the final products I will discuss below. The important part is that they are clear and have the correct overlap or your end products will suffer.
This is important so it gets it’s own paragraph. Still images are the backbone of mapping with your drone. Learn how to collect them and collect them well.
Videos
The videos you record are a lot more straight forward. Mostly you just need to be a good pilot to collect cinematic shots. You can then slice and dice them any way that you want to suit your needs. This website focuses on still images and what you can do with them. However aerial videos are a powerful tool in themselves and absolutely they are data.
What are the final products?
The final products of the data you collect with your drone are where this whole thing gets interesting. They are created using the process of Photogrammetry. Photogrammetry is a very complicated mathematical science, but don’t worry you don’t need to know it. I don’t even know the math behind the process. The important part is the software knows it, so you need to know the software. If you want to learn more Science Direct has a great series of chapters on the topic located here. Photogrammetry gives us all of these great products and that is the reason we need to pay attention to all of the different variables in our pictures especially the overlap.
Photogrammetry is the art, science, and technology of obtaining reliable information about physical objects and the environment through processes of recording, measuring, and interpreting photographic images and patterns of recorded radiant electromagnetic energy and other phenomena
Wolf and Dewitt, 2000 and McGlone, 2004
So now that all the tough/boring/hard to understand stuff is behind us… What are the end products for the data that you’ve collected? The first one is so obvious that you almost forget it. The Videos! Those don’t need any special processing except to edit them into a usable product. The rest are a little trickier but I would say much cooler. Once you run your raw data through some photogrammetry software you can expect orthophotos, elevation models and 3D products such as point clouds and textured mesh.
Orthophotos
Orthophotos are basically stitched together images that are orthorectified. That is another one of those complicated processes that the software knows for you. If you want to learn more click on the link in the first sentence. However here is a quote from the same article explaining the process – “the process of removing image distortions or displacements caused by sensor tilt and topographic relief”. Basically what you are doing is creating a Google Maps image with insanely good resolution. I’ve seen the resolution get down to under 1cm/pixel. However remember with great resolution comes small collection areas. You can typically only map 20-100 acres at a time.
Elevation Models
Elevation models are a raster dataset that gives each pixel a value that represents the elevation at that point in space. Elevations can be a little tricky. Only licensed surveyors can legally represent their elevation models as true and accurate. If you are not a licensed surveyor do not guarantee your products as true and accurate. At most if you are unlicensed you can say your data is for visual purposes or rough estimation only.
With that legal stuff behind us, what can you do with an elevation model? Well you can show the elevation across the entire area that you collected data. You can use this to generate contour maps of the area for things like drainage and slope calculations. You can also create Hillshades which are a great way to visually see the landscape you’ve just collected.
3D Products (Point Clouds and Textured Mesh)
The 3D products you get back can be some of the coolest final results of your whole data collection. The point cloud is the recreation of all the data you collected represented as tons and tons of points. You could also think of them like pixels only on a 3D plain. Point clouds are one of the greatest things you can get out of photogrammetry. They are great for measurements and visualizing spatial relationships as well as creating 3D modules of an area of interest.
The 3D mesh is a derivative of the point cloud. It like it says it is, it meshes the point cloud together by connecting points and making surfaces with them. Photogrammetry creates a 3D visual of the point cloud with out the spaces in between all the “pixels”. It can be a great looking 3D product but it’s more useful as a visual and less useful for measuring things. However I find that the point cloud in most cases looks better and is a more accurate representation of the ground, both visually and analysis wise.
So what’s next?
So what IS next? This article is a great introduction to the whole E2 Spatial ecosystem. I’ve really opened a can of worms here and you could go down many trails. Next up we need to learn about Photogrametry software, Data Collection and many more topics. But above all we need to remember that we are professional pilots. Make sure while you’re collecting data that you are conducting yourself as a professional.
All raw data was downloaded from DroneMapper and processed by myself
Please visit my Resources page for more resources on the topic of Drone Data.
