Imagine your drone suddenly running low on power - a long way from home, doing it's auto waypoint flying thing.
You have to decide how, where and when to set it down very quickly!
I've been playing with my homebrew Automatic Landing Finder (ALF) that uses a photo to determine where a suitable landing spot would be. The algorithm takes things such as size, visibility etc. into account when determining what spot to pick. The system picks the best candidate among possibly thousands.
The idea is to have a camera pointing downward taking snapshots at regular intervals. Should things go haywire, the system will analyse the last frame where the craft was till stable, and try to find the best landing spot.
Eventually this system will be installed on the next version of the USAV (Unmanned Social Autonomous Vehicle) but for now it's being tested with "fake" images from google earth.
If you had to land anywhere on the picture below, where would it be?
Feeding the image above though the ALF reveals the places that ALF considers as being a more or less good spot to land. The resulting image is greyscale because this is a part of the process. Yellow squares are suitable candidates, and the red square is the chosen optimum landing site. Observe how dark spots (where poor visibility prevents a proper condition estimation and vegitated areas are (mostly) avoided. Did you chose the same spot?
How fast did you determine a proper place to land? The computer took roughly 842 ms to find a good spot.
Below is the same result with all candidates removed, only showing the optimum landing site.
I'd love to get some feedback on what you flying guys out there think is a good emergency landing spot in general as well as what is a bad one! :)
Best regards
Jesper Andersen
Comments
yes, precisely the same place, and it took me about 5 seconds, but I was a bit distracted. If your code can get it in ~850ms that's pretty good!
It's worth picking the brains of experienced glider pilots who need to identify landing surfaces reliably for outlandings.
Wow this is really thinking "outside the box". It would be interesting (although perhaps impractical). to imagine whether any of the "big players" (Lockheed, Boeing, BAE systems etc) have thought of something along this line?
Its definitely something worth persuing. Well done.
My memory about IR vs Visual is really shaky, and this might require a secondary sensor - or a camera capable of seeing both visual and IR - but isn't there a differential between water in visual and water in IR that you could trigger off of?
Ya there are actually a lot of factors that go into what color a body of water happens to appear at any given time. Particulate matter from runoff can make it all brown. Tannis can make it brown to black. Clean fresh water can vary depending on the depth. It could be more of a reflection of the surroundings in shallow water, or more blue for deeper water. It may not be be practical to detect and manage all of that.
Very clever. I was briefly involved in simple contrast-based machine vision development in the 90's (tool identification and selection) but your addition of color to an algorithm is brilliant. Can't wait to see where this goes.
I think you did it perfect, the spot is the most open and it is not in the middle of the road
I was going to suggest avoiding roads and pathways but that might overcomplicate the process and cause errors, maybe keep a few backup spots just in case the approach changes
This is amazing stuff. This is exactly the kind of blog posting that restores my faith in this project after fighting in the clone wars for the past week. ;)
So without looking at the location your algorithm picked, I chose this location:
I chose this, probably for exactly the same reason your algorithm did, but I decided to try to get just a bit farther away from the trees, with the risk that site is potentially just a bit rougher. It took me... a few seconds to make the choice. I think you're doing very well!
The assumption that water is blue is probably not great. Where I live, most natural water is green (algae). Sometimes it's brown (tannins).
Very interesting Felixrising - I wonder how a NIR cam would do?
You can see the bands that h20 give off ... using a TIR cam... http://www.geog.ucsb.edu/~jeff/115a/remote_sensing/thermal/thermali... various bands here: http://www.geog.ucsb.edu/~jeff/115a/remote_sensing/fig1_5spectralre...
-
1
-
2
-
3
of 3 Next