Hi all,
May I begin by saying what a valuable forum this is, although this may be my first post I’ve been following various threads for a while and there is nowhere else on the internet with such a mix of high level technical expertise and accessible knowledge base.
I’m a graduate student working on a project in conjunction with the US Army Corps of Engineers whereby we plan to monitor change in fluvial geomorphology and sediment conditions in a section of the North Fork Toutle River, Mt St Helens. I would greatly appreciate some of your insight into our plan to capture high resolution, low altitude, vertical aerial imagery in order to create accurate orthoimages and possibly a DSM. I have a background in geomatics and GIS, but not electronic engineering or UAS’s so my apologies if some of my questions seem obvious.
The area of interest can be seen in the image below, the yellow bar being 500m (UTM).
The plan is to do complete repeated fly-overs with a UAS to gain decent resolution imagery and generate an orthophotomap. We have access to good GNSS kit and will be running a few dGPS ground surveys, hence we plan to establish a number of ground targets to act as high order control points on the imagery. Our equipment budget for this element of the project is extremely limited (~$200).
My first question is what platform would be best suited for this mapping application? I like the idea of a multi-rotor system. From what I’ve read on here some form of quadrocopter would most likely be ideal due to stability and ease of flying. Note that no ground station would be required.
My second question relates to GPS/IMU units on the UAS. I’m assuming that the presence of ground control targets may negate the need for GPS on the system? Of course, GPS would be a benefit in terms of autonomous flight lines and tracking, but this is not essential? However, I’m not sure whether IMU is needed for interior orientation of the image block (Ωφκ?). If so, how easy would adding an APM 2.5 (+GPS) to a low end quadrocopter be?
Finally, in terms of a camera. We’ve got access to a D100 but I’m assuming this would probably be too bulky to mount. Would a simple 10M point and shoot be suitable? Unfortunately I can’t find much information on triggering devices, I don’t think the system need to have a solenoid mechanism or anything too complex; hence would it be possible to to simply set up some sort of trigger timer on the camera?
I’m very grateful for everyone’s time and consideration and look forward to the discussion.
Replies
I would use a flying wing from Rob with a Canon SX230 (the number might change depending on which country you are in, its the one with GPS) I would fly from that road on the right trigger the camera with CHDK and then I would use the dronemapper.com to process them. All with an APM 2.6 keeping it on the straight and level. If I wanted to be really fancy I would add some 433 telemetry modules and let them speak to an Android tablet so I can keep an eye on the machine in flight.
Of course your flight there would be beyond line of sight so not strictly legal. The other problem is the budget you won't get that for $200 and you absolutely won't get a multirotor for that price.
So if it must be aerial images that probably leaves you with KAP, kite aerial photography. Its a black art but absolutely delightful when you get it right. It has the advantage when the wind is there of persistence. Make Google your friend for that and take the time to make a picavet its not hard.
Good luck.