Author Topic: Breaklines and mass points  (Read 2849 times)


  • Newbie
  • *
  • Posts: 32
    • View Profile
Breaklines and mass points
« on: January 24, 2019, 09:40:29 PM »
Hi everyone,

What would you say is the most precise and efficient way  of creating breaklines and mass points from UAV data sets. I normally don't do this for every project, only those were roads, creeks, rivers, mine pits and so are present, this way generating a more accurate TIN vs just working with the DSM and DTM generated. What I'm doing right now, with the resources I have, is draw polylines directly on the point cloud using agisoft and exporting this to arcmap were I generate a TIN based on those 3D polylines. This is time consuming and in some cases not as accurate as using the stereo pairs, which isn't an option with UAV data because of the amount of images generated in a flight.

I found two software that show how to create breaklines from point clouds, but really don't know how they would work in certain conditions where lots of vegetation is present. One of the is Carlson Software and the other is virtual surveyor. If anyone has worked with any of this please provide some feedback or if you have a much better aproach it would be interesting to hear about it.

I must also add that I'm not a photogrammetrist, my background is more based on GIS software but with no photogrammetry training, everything I've learned about photogrammetry has been through experience.




  • Full Member
  • ***
  • Posts: 167
    • View Profile
Re: Breaklines and mass points
« Reply #1 on: January 24, 2019, 10:33:19 PM »
"which isn't an option with UAV data because of the amount of images generated in a flight"

My dear friend you are completely wrong, if you want to extract quality lines in the areas that mention this is the way, it is a very common problem nowadays, some people want to work on this, but this is photogrammetry, if you need precision and jobs of quality in these areas you must use stereo pairs, but you need software and appropriate hardware and above all a lot of training and training, of course if you really want to do something of quality in these complex areas where automatic correlation processes are not going to work, keep in mind that this It is not lidar, so the pointclouds look similar. Now if you need simple mosaics then you can do what most people do.

Independent of the number of images there are appropriate software for the task you need, but it requires training and hardware, it is not as simple as making automatic mosaics.


  • Hero Member
  • *****
  • Posts: 505
    • View Profile
Re: Breaklines and mass points
« Reply #2 on: January 25, 2019, 04:50:00 AM »
I have workes with virtual surveyor, and Must say semi-automated breaklines extraction is the best approach I've seen to the date. It works really well  in naked areas and due to its hability to allow user input, it is rather usable even with trees or some low vegetation. The only drawback it is that VS is designed to work in 2.5D mode (all surfaces have Z positive normals), not allowing true 3D models to be "virtually surveyed".