Linear mapping project

Hello all,

I have a potential opportunity for drone documentation of a 22 mile pipeline route along an existing high voltage power line. All of my mapping projects have be traditional grid type sites so I have no experience with this type of project. I have also flown a number of times in the vicinity of high voltage lines but never along them for any extensive length. I am certain that I would be flying well above them as high resolution images is not a critical requirement.

No survey control is required as the project requirement is simply to document existing conditions along the route for inclusion in a pre-bid package. The client initially asked for video along the length of the route (they know nothing about drone regs, line of site, flight range, etc) but it seems to me that a linear mapping route would be much more practical and useful.

My question is, never having used the linear route function in DD, would I need to plan the flights in line of sight segments or does the linear route option in DD have some method of establishing the entire route and them breaking it up the segments to allow for daisy chaining of takeoff/landing points along the route? Many other questions to be answered such as access points and ability to travel by truck along the route but, obviously I will have to answer those by research and reconnaissance. Most of the route is in a rural area so following the drone in a vehicle is a potential option if conditions allow vehicular travel along the route.

Thanks in advance for any input.

Steve

1 Like

Are you just needing images or are you looking for an accurate stitched map? Either way I don’t think DroneDeploy is good for long linear missions. You might checkout DJI Terra if you are most interested in automatically splitting up a very long project. I currently have a 10-mile roadway that was really easy to split manually so I don’t know that it is that big of a deal.

Hey Michael,

Thanks for the quick reply. My initial idea is to stitch a single row of photos along the route into one long map or as few a set of segments as possible to be viewable by potential bidders. Survey grade data is not an issues as the route has already been surveyed by traditional methods. My only requirement would be that the photos stitch together in a reasonable manner so the distortion doesn’t make it un-viewable. If accumulated stretch becomes a problem (say after several miles, hopefully) then I would break into individual segments for processing. So in short, the internal drone GPS should be good enough. I just want the segment to be positionally accurate enough to overlay on a GIS background map.

The client actually asked for a continuous video like a low flying helicopter would shoot by just flying from one end of the route to the other without interruption. That makes since at first blush, but after more consideration I see problems: 1) geolocation - if you pause the video to look at something, how do you know where your looking, this is flat rural country and everything tends to look the same from the air; 2) indexing - I have fast forwarded through videos before trying to find what I was looking for. It’s hit or miss at best.

Seems a stitched photo map would solve all of these problems, especially since survey grade measurements are not a criteria. For flight control, all that I really need is a method to fly along the route an snap photos at regular intervals (assuming that a single line of photos is stitchable - two rows would complicate things but might be necessary). The second requirement would be a method of easily geo-locating where the last leg left off if/when I have to drive around an terrain obstacle to continue flight control from the other side.

Again, any input greatly appreciated. I will give DJI Terrain a look.

Steve

A single line of photos could cause issues because of the lack of tie-points available. Even if you flew with and 80% frontlap you are only going to get 4-5 images per pixel which will stitch, but the longer the flight the more warping you will get.

I naturally broke mine up into 1-mile sections because that is a nice manageable segment length as far as documentation goes, but it is also in accordance with line of site unless there are obstructions which it doesn’t sound like your scenario will have. I can see and navigate the drone up to 3000ft on good days so a mile section with me stationed in the middle made sense. This is what DD is not good at. It has to start at one end or the other so stationing in the middle causes ALLOT of battery waste to start and from end on that length of a mission.

On this project I used my Yuneec H520 and DataPilot, but it would be the same thing with a P4P and Litchi except for the fact that Litchi does not automatically break it down or define 2+ lanes off of a single centerline. It really didn’t take very long to put together so I don’t mind the little extra work to get exactly what I want. Additionally I am flying the site bi-weekly so the flight planning was worth it.

I had two lanes with 75/60 overlaps which gave me 8 images per pixel and it stitched well, but without GCP’s there was still slight warping. Photogrammetry in general has an issue with long linear missions. In a 2D ortho it was not that noticeable until you started measuring long distances.

I chose to use GCP’s and made sure that the ends of each mile shared a point. We did need grade, but more importantly it made sure that each mile would align to the rest so that one master ortho could be made in QGIS by combining the orthos of every flight. This is very easy to do.

Thanks Michael,

This is great info. We will be discussing with the client today. As always, time is a factor. If they are interested in pursuing I plan on doing recon and test flying/processing a short segment this weekend.

BTW, have you had any problems with electrical interference? I estimate the towers to be 80 to 100’ tall and I expect to be flying at 200’ so.

Thanks again

Steve

Sounds good. I haven’t had any trouble with flight, but lines have caused a little decrease in video range when they are high-voltage and on the flight path. This didn’t happen every time though. If you are flying the two-lane method that should keep you on either side of it and it shouldn’t be a problem.

Just as additional information one mile stationed in the middle took about eight minutes to fly so you should be able to get about 2 mi on one battery and can calculate from there.

I look forward to hearing more!