Right now it does not seem possible to design a mapping area that has curves inward (in the direction of the center of the polygon). This means for example that when mapping a river that meanders a lot, you'll end up mapping way more area than necessary. For example, you can not design a flight area in the shape of an L or Z.
Besides mapping more area than necessary, it makes it more difficult to map areas in rugged terrain, since you can't avoid high obstacles by taking "bites" out of the mapping area. Currently the only way to work around that is flying (way) higher, which might not be ideal or even illegal. Dividing the flight plan into 10 or more flight plans doesn't sounds very practical to me either.
We are considering the addition of a linear layout system for mapping rivers and pipelines. DJI changes the SDK so frequently it seems like we spend most of our time just playing catch up... We do not plan on adding any sort of arbitrary concave polygon support. L and Z shaped surveys would be covered, crazy zig zag shapes like a sunburst will not.
댓글을 남기려면 로그인하세요.
I also have difficulty problem
Hi
This could also be useful when mapping roads that curves like a river.
For us it would be perfect if we can draw a polygon that follows the road and then the Inspire flies in straight lines within the polygon.
Is this possible?
Regards
Andreas
NCC Construction Sweden
I have a similar request. L or Z shaped surveys would be great! Has there been any progress on this since January?
I wish to survey a generally square property for one owner of which a small portion in the top right corner belongs to another owner. Currently, Map Pilot will not permit me to survey all of the original owner's property without overflying the other owner's portion of the area.
Just curious how others are handling this. Thx
We recommend using multiple flights to achieve complex shapes. An L would take two flights, for example, and a zig zag would take 3.
You can always omit images of areas you don't want to include.
Don't black them out. You would be better off processing the whole area and then masking it using other software is it is that sensitive. It can be pretty hard to control where the edges land and we can guarantee where the borders will be.
Of course! I've just been overthinking this.
So, I just do two separate flights to cover the L-shape. Then upload all the images from both sets of images as a single batch to create an L-shaped map. That's easy enough.
Cheers!
I'm a newbie at this, so excuse me if this is a dumb question. I want to create an L-shaped map using two separate flight grids.
I just upload the two batches of photos together all at once to create a single L-shaped map, correct?
Does it make any difference if the photos are out of sequence? I mean, do they have to be uploaded in any sort of sequential continuity, or will your software automatically stitch the two batches together as long as there is sufficient overlap in the photo coverage between the two batches regardless of the order in which they're uploaded?
I'm using a Phantom-3 Pro, so I guess the images are all geo-tagged with EXIF data. I'm guessing that helps your software stitch them together regardless of the order in which they're uploaded.
I'm also planning to map a ~30 acre partially forested lot, but am worried about being able to maintain sufficient line-of-sight transmission signal between me and the drone as I'll be down in the forest. Can I break the job down into four smaller quadrant flights, then upload all the photos from the four batches at once into a single map?
This is sooooo cool! The more I read, the more excited I'm getting about this. I'm hoping to get out this week to do some experimenting on a smaller plot. That will either answer most of my questions, or probably generate many more questions...
Sequence does not matter. You can upload images from different flights all at the same time as long as they are from the same takeoff location (otherwise the height offsets will get weird).
Hmmm, same takeoff location? Good point as it ensures a consistent flying altitude.
That might be a problem though as I was hoping to take off from four different locations, one in each quadrant, to ensure the drone always stays within sight during flight. I could take off from one central point, equidistant to each of the four quadrants, or even do the whole thing in a single flight. But I'm worried I might lose the signal, because of the trees, when the drone reaches the far corners of each quadrant.
If it's a question of matching height offsets for each of the four flights, how about if I pick four convenient takeoff locations and note each of their ground level elevations on Google Earth, then adjust each of the four grid AGL altitudes accordingly so that the entire grid together is on one consistent ASL altitude plane?
Would that work? Assuming the Google Earth indicated elevations are accurate enough...
In other news, I flew two test patterns this afternoon. It works brilliantly! Yippee! Gotta run now, but will try uploading later this evening. :)
If you were to scout the takeoff locations ahead of time and get their elevations you could ensure that all of the images will be taken at the same elevation, which is the important part.
For Example:
Target AGL from takeoff location A is 100m and has an elevation of 500m above sea level as found by Google Earth.
If takeoff location B was at an elevation of 520m above sea level you would set the flight altitude to 80 m. This might make the overlap a little different in those locations but if the terrain tracks the takeoff elevation heights it should actually turn out pretty nicely.
Excellent! That's exactly what I was thinking re: adjusting flight altitudes relative to takeoff height. The area I'll be mapping is roughly 400m x 400m square, sloping down to a lakeshore with a difference in elevation back to front of only about 10m. It can easily be covered with a single flight. It's just that I'll be standing on the ground, sometimes with tall trees blocking my view of the drone. I plan to first do some test flights to check transmission quality and, if good, then a single flight grid will cover it. I just prefer to have a Plan-B in place in case I'm forced instead to do multiple flights from several takeoff locations.
I have one other question though. (By the way, you guys are GREAT! Thanks for your patience.) I'm just now in the process of uploading my first smaller test-map. It asked me to include one photo at ground level from my takeoff location so it will more accurately be able to fix the elevation of the resulting map. That's cool. The iOS Map Pilot App snapped one final photo as the drone was landing and I wondered what that one was for. One small issue though -- I always hand-catch the drone when landing, so that final photo is actually about 6' above ground level. :)
Anyway, my question is, if I do use four different takeoff locations to stitch together my lakeside map, will it matter which one (or all?) of the four locations I use for that ground level image?
It won't matter as long as all the flying is done at the same height with respect to any of the takeoff locations.
400m x 400m isn't that big. Take off from the highest side of it in a clearing if you can, set the overlap at 70% and go. If it is dropping images you can always start over the other way.
Starting from the high side of the lot was my original plan, but that puts the lakeshore the furthest away with lots of trees in between. I'd prefer to finding a clearing somewhere more central in the lot. It's about a two-hour drive from my place, so I wanted to have a firm plan (and backup plan) in place before setting out. I'm probably worrying about nothing, but I'm a worrier by nature.
One final question and then I'll stop bugging you. :) Somewhere in the many posts I've waded through in the FAQ and Forum, it was suggested that for better coverage and overlap the map grid should be flown twice, with the grid rotated 90 degrees. Is that still recommended?
That is more of a recommendation for doing full 3D models where there are lots of vertical or overhanging faces that you would like to get more detail on. For mapping, it isn't really needed.
This raises a question for me. Does Map Pilot use above sea level or above ground level when flying missions? If it is sea level, then the height over any particular part of a mission will vary from one spot to another as the terrain changes. If it is above ground level, then as the terrain changes, so will the drone height.
In either case, I'm not sure why you would need to adjust the height offsets. 100 meters above sea level will always be 100 meters above sea level regardless of the take off point, and 100 meters above ground level will always be 100 meters above ground level for the same reason. Unless I'm misunderstanding, and the mission height is based on above ground level at the point of take off, then remains at that altitude regardless of the terrain below.
Not that we actually recommend doing it but concave survey area layouts are available in 3.1.0 which was released today.
