Written by Greg Barnes
Here at 2cofly, we have done many photogrammetry and LiDAR projects both here on Guam and in Palau. Our clients love our work, and we love aiding them in the success of their businesses.
Depending on the outcome that the client needs, we will decide which sensor would be better for us to use — an RGB camera for photogrammetry, or a LiDAR sensor which uses lasers. In this relatively short article, I’ll give a pretty short, easy comparison of each.
Photogrammetry. When you take a normal photo, you’re basically taking a 3D object, like a real-life person or a house or whatnot, and converting the image of that object into 2D — into a picture. So 3D to 2D. With photogrammetry, you’re taking lots of overlapping 2D pictures and basically recreating that 3D object out of it. So let’s say you’re flying your drone over a house. The drone flies above the house, and as it moves around, it takes lots of photos of the house from different places. Later, you put the images into some specialized software that analyzes all the different photos and does some very complex mathematical equations and analysis on the photos.
When this is done, the outcome can be an orthomosaic map (read all about what that is here), a 3D colorized point cloud, or even a 3D model of that object. (For most orthomosaic maps, the drone will fly with the camera facing straight down. For 3D models, you’ll want to have the camera at a slight angle.)
One thing to keep in mind: What the camera sees is what you get. There is no X-ray vision with a camera — we can see what’s right out in front, but not what’s behind or underneath something. (Often, this is a good thing, especially when someone very ugly like Aunt Gertrude or Hilda is trying to hide behind someone else because they don’t want to be seen. Yeah… we don’t want to see you either.) A similar consideration under the “what the camera sees is what you get” category, is that obviously for photogrammetry, the camera needs to pick up visible light. Because of this, photogrammetry missions would only work during the day. The dark of night obviously isn’t good for photos.
Photogrammetry can be done with photos taken from pretty much any drone, but some cameras are far superior to others. With most drone cameras, by nature of the camera shutter, the accuracy of your outcome may lack. And that’s because this is done with a camera. And as I said, some cameras are better than others.
You know the basics of how cameras work: light enters the lens and yada yada. Cameras are passive sensors — that means that information about light is coming into the camera. (The camera is not shooting anything out — so passive sensor — unless it is actually a gun disguised as a camera a la old cartoons.)
LiDAR (Light Detection and Ranging), on the other hand, is not a passive sensor; it is an
active sensor. This means that yes, the LiDAR sensor actually IS shooting stuff out. In the case of LiDAR, the sensor is shooting out laser pulses. These laser pulses are invisible to the human eye and safe to the human eye. These beams of light travel through the air at, well, the speed of light (………yes, one would hope). When that pulse of light bounces back into the LiDAR sensor, it is recorded how long it took for that beam to shoot out and return, thus measuring the distance. (If one beam takes longer than another to return, that means that the distance to that object is further. It’s like an echo. A fast echo = a small canyon or cave. A long echo = a larger canyon or cave.)
Just like you can do with photogrammetry, LiDAR creates a 3D point cloud. If your LiDAR sensor includes has an RGB camera, that LiDAR point cloud can be colorized, just like with photogrammetry.
One big difference between LiDAR and photogrammetry is that, while you need visible light for photogrammetry to work, you actually don’t need that for LiDAR. Since the LiDAR sensor shoots out bursts of laser light, no external light source is needed. So, if you needed, LiDAR missions could even be conducted in the black of night. Of course, you would still perform those missions in the daytime if you wanted your point cloud to be colorized. But if the point cloud is the only important thing — not necessarily a colorized one — then sure, you could go ahead and fly at night.
Another a big difference between LiDAR and photogrammetry: LiDAR laser beams can actually shoot through certain thin objects (such as leaves or thin branches). Remember what I said above: with a camera, what you see is what you get. Well, with LiDAR, you get more than just what you see. Here’s a fun scientific fact: As laser pulses pass through one object and on to the next object hidden underneath, the laser pulse actually freaks out and says: “Aunt Gertrude? Hilda?!”
So let's say you are a construction
company and you need a topographic map. Is the land cleared of all vegetation? If so, sure, we can make a topographic map for you using photogrammetry or LiDAR. However, if there are trees on the property and you need to see the layout of the ground underneath the trees, we’d need to use LiDAR for you. (We’ve done that in many places, including both times we went to Palau. Both times, we were scanning areas of jungle. If you’d like, you can read all about our first trip and second trip to Palau.)
That’s it for now. I hope you learned something, I hope you enjoyed, and I certainly hope you don’t tell your Aunt Gertrude that I talked about her like that. (Hilda, on the other hand, I don’t care if she finds out, I reeeeally don’t.)
blog written by
Greg Barnes
Operations Manager
(671) 689-7939
Awesome article Greg. It goes to show why drones with swappable payloads are so capable. Putting on a LiDAR sensor and capturing what lies beneath the tree line is nothing short of amazing. Love seeing this kind of stuff and how drones are becoming an awesome tool