Anybody have any tips for flying irrigation rigs in fields?
Auto flight? Videos vs pictures? Flying while rig is operational, …
1 Like
What kind? Pivot, linear? How big?
Pi iOS is the primary rig in this area. Most are 1,000 - 2000 feet long.
Here’s a previous discussion on the forum regarding pivot rigs. I think the trouble you’ll find is that if you are trying to capture the rig itself (and not necessarily the field) then it is going to be tough. You would have to fly super-low to capture any detail so it would be very time consuming.
Thanks. I’m thinking about inspection pictures. Very low level, and not trying to capture the field with it.
Right now thinking video from about 50 feet. Wondering if there is a way to capture a series of photos automatically instead of video. Fly down one and side back up the other.
Another set of experiments to plan out.
Thanks for the reference.
Gotcha. Inspection photos are super easy to do in manual flight. If you have a 4k monitor I would video in 4k and then take screen clips of the video. They will then be 4K pictures, but they won’t be geo-referenced. Doesn’t sound like you need that for inspection photos, but if you did need them you could give them GPS coordinates in Geosetter.
TVOSHELL: I don’t know whether or not you’ve considered flying gently curved, convergent, non-traditional (non-linear/non-parallel) flight lines before to a) provide a more diverse view perspective and b) help mitigate the well known Structure-from-Motion doming [elevation] error. The example provided illustrates such a curved flight line mission. The flight lines shown are log-spiral curves and no two segments of any flight line is parallel to any other flight line segment. A strong advantage to you might be that each flight line (trajectory) has its own constant tangent angle such that the camera does not need to be moved at all in order to focus on the central portion of the target area when flying any single flight line - note the yellow pins (target focus) placed at each end of the mission plan. You could use high-res
still frame photos or frames from a video.
1 Like
Hey James! Glad to see this come back around. I think we have determined at this point that Thomas is doing inspections and primarily interested in imagery for documentation and not mapping. I’m not sure where this method would even matter if it was mapping, but I think you should start a refreshed thread to provide some updates on this theory.
Michael: Thanks for your response and suggestion. I’ll put some thought into it.
I am currently using the curved flight for a) mapping large areas and b) inspecting electric power transmission structures and corridors. Only high resolution frame imagery (oblique) is used.
I’m using manned aircraft (helicopters and fixed-wing) and long-endurance fixed-wing drones to capture the images. It’s a good business.
Jim Dow
Great suggestions. For a pivot irrigation system I need to fly around 50 to 100 ft altitude with detailed photos of the full system, about 1000 - 2000 ft long. Structure integrity and functionality (are the valves spraying correctly) are of importance. These are existing structures that are difficult to get to once the crop is mature. What I think I need is a constant camera angle and position and a flight path along a straight line. Video is fine but grabbing single frames doesn’t provide the detail a single photo will. The more detail the better on these rigs. Think flying a tower that is lying on its side. Flying a central POI won’t provide the results needed.
You use that flight pattern for corridors? Seems like that would be hugely inefficient and not a case for the “non-linear” flight. I do allot of pipeline and roadway and this would be a nightmare.
Seems like the best flight plan for this case would be 1 nadir and obliques from each side perpendicular to the rig. This could be done pretty easily with Litchi.
Not really. Like me, you would choose to only use two or three curved flight lines.
If you are using 2 or 3 straight parallel flight lines on pipeline and road corridors, I can’t see how you are not seeing the SfM doming issue - seriously.
Jim
When we are doing visual inspections it really doesn’t matter.
I’ve been reviewing Litchi and would agree. More training time ahead.
Would you mind explaining what the SfM doming is?
In my opinion this has all but been mitigated over the years with progress in drone and processing technologies. I had it happen less than a handful of times back when I flew and Phantom 3 Pro 4 years ago and have not seen anything like it since. This include miles of roadway and pipeline with perfectly parallel flight lines.
Here’s an abstract from one of the lead studies, but you can find more on the forum if you search for doming. Let’s not muddy up this thread with that discussion.
High resolution digital elevation models (DEMs) are increasingly produced from photographs acquired with consumer cameras, both from the ground and from unmanned aerial vehicles (UAVs). However, although such DEMs may achieve centimetric detail, they can also display systematic broad‐scale error that restricts their wider use. Such errors which, in typical UAV data are expressed as a vertical ‘doming’ of the surface, result from a combination of near‐parallel imaging directions and inaccurate correction of radial lens distortion. Using simulations of multi‐image networks with near‐parallel viewing directions, we show that enabling camera self‐calibration as part of the bundle adjustment process inherently leads to erroneous radial distortion estimates and associated DEM error. This effect is relevant whether a traditional photogrammetric or newer structure‐from‐motion (SfM) approach is used, but errors are expected to be more pronounced in SfM‐based DEMs, for which use of control and check point measurements are typically more limited. Systematic DEM error can be significantly reduced by the additional capture and inclusion of oblique images in the image network; we provide practical flight plan solutions for fixed wing or rotor‐based UAVs that, in the absence of control points, can reduce DEM error by up to two orders of magnitude. The magnitude of doming error shows a linear relationship with radial distortion and we show how characterization of this relationship allows an improved distortion estimate and, hence, existing datasets to be optimally reprocessed. Although focussed on UAV surveying, our results are also relevant to ground‐based image capture. © 2014 The Authors. Earth Surface Processes and Landforms published by John Wiley & Sons Ltd.
tvoshel: SfM (Structure-from-Motion) is an image processing technique (commercial software such as PIX4D, AgiSoft, OneButton, etc.) used to process a collection of image frames to produce a) digital elevation model, b) orthographic photo, and c) 3D model of the photographed scene. You may want to look into this technique. Other than DroneDeploy, I would recommend looking at AgiSoft.
For your application (structure inspection) you do not seem to need SfM type processing. I would recommend you use 3 straight and parallel flight lines: a) one vertically over your structure [right over the centerline of your structure] with your camera set at nadir and b) two additional flight lines, one on either side of the structure and parallel to it with the camera set to an oblique angle of 20 to 30 degrees up from nadir and perpendicular to the structure centerline. Taking overlapping photos, you will be able to see 3 sides of the structure. Each of your photographs should be georeferenced to the position of your drone at the time the photo is taken. Be certain that the georeferencing information is written to the EXIF metadata of the photo.
Take a look at an inspection software SCOPITO which will allow you to view all of your georeferenced photos relative to each other (all at the same time) and annotate each photo (one at a time) with any issue you encounter. The SCOPITO software will produce a nice report for you - for further reference.
I use the SCOPITO software quite a lot for inspecting electric power transmission structures as well as the corridor.
I hope I have been helpful. At this time I don’t think you would benefit from the curved flight line mission plans. However, the SfM doming [elevation] issue still exists (very much so) in the commercial software environment; so you might consider using curved flight lines in the future if your applications warrant its use. I use curved flight lines quite a bit using piloted fixed-wing and helicopter aircraft as well as fixed-wing long-endurance drones - with continuing success. I’m attaching some examples of curved flight line missions for your review.
Thank you. This is very helpful and I hope others have gained information from it as well.
1 Like
I am currently using the curved flight for a) mapping large areas and b) inspecting electric power transmission structures and corridors. Only high resolution frame imagery (oblique) is used.
2 Likes