Author Topic: Jagged profile from surface created from points...HELP  (Read 5233 times)

0 Members and 1 Guest are viewing this topic.

alanjt

  • Needs a day job
  • Posts: 5352
  • Standby for witty remark...
Jagged profile from surface created from points...HELP
« on: September 18, 2014, 09:52:14 PM »
I have a group of points that I'm trying to create a surface from, but no matter what I do, when I run a profile down the area that is a road (has breaklines and a pline on the zero I've been using to create a quick profile), I keep getting a very jagged profile. I keep trying to simplify the surface or weed points, but nothing I do will fix the problem. Any thoughts would be greatly appreciated.
Civil 3D 2019 ~ Windohz 7 64bit
Dropbox

Jeff_M

  • King Gator
  • Posts: 4087
  • C3D user & customizer
Re: Jagged profile from surface created from points...HELP
« Reply #1 on: September 19, 2014, 02:11:37 AM »
Looks like the profile is just doing what the points defining the surface are telling it to do. The best you can do is to be sure the polyline/alignment follows one of the breakline plines so it doesn't cross TIN lines

mjfarrell

  • Seagull
  • Posts: 14444
  • Every Student their own Lesson
Re: Jagged profile from surface created from points...HELP
« Reply #2 on: September 19, 2014, 09:13:16 AM »
I've applied some surface smoothing Kriging Method....it looks better however that isn't the solution per se.

The issue is really did the survey crew collect data with a high enough resolution for the intended purpose?
This is something that is a learning curve for some, to get enough data to accurately represent the surface.
With a good sample interval one can almost do without breaklines.
Be your Best


Michael Farrell
http://primeservicesglobal.com/

alanjt

  • Needs a day job
  • Posts: 5352
  • Standby for witty remark...
Re: Jagged profile from surface created from points...HELP
« Reply #3 on: September 19, 2014, 01:36:23 PM »
Looks like the profile is just doing what the points defining the surface are telling it to do. The best you can do is to be sure the polyline/alignment follows one of the breakline plines so it doesn't cross TIN lines
Yeah, that's what I figured, but I was hoping there was a way to smooth the data out a bit so the profile didn't look so horrible. The area in question is a railroad, so I'm quite sure the ground doesn't actually jog that much.

I've applied some surface smoothing Kriging Method....it looks better however that isn't the solution per se.

The issue is really did the survey crew collect data with a high enough resolution for the intended purpose?
This is something that is a learning curve for some, to get enough data to accurately represent the surface.
With a good sample interval one can almost do without breaklines.
This is flown LIDAR data. The drawing I posted is a blocked-out portion of the larger area that was sent to me. The only thing I edited, was putting a point at each breakline vertex, flattening the polylines, and using them as proximity breaklines. The only reason I did this was in the hopes that, when I ran a 'simplify surface', it wouldn't be bound by elevated breaklines.

It's a pretty tight grid through there, but it's just really jagged, and I can't understand how LIDAR data would yield such odd results.

Another oddity was in the elevated contours they provided. When I created a surface from them, the c3d contours matched the polylines - as they should, but when I would pick points to check an elevation, it would return strange results. If I picked between 100 and 101 contours, even in the exact middle of the two contours, it would give 100.01, but when I got right up to the 101, it would jump from 100.02 to 100.9.

I just need to be able to show a good profile of the existing area and I can't get it to smooth itself out.
Civil 3D 2019 ~ Windohz 7 64bit
Dropbox

mjfarrell

  • Seagull
  • Posts: 14444
  • Every Student their own Lesson
Re: Jagged profile from surface created from points...HELP
« Reply #4 on: September 19, 2014, 02:33:09 PM »
I suspect you were not given the entire raw data file for the lidar.

I've used some here and there and it rarely needs breaklines to 'enhance' the surface that is developed from the point cloud(s).

I see no value in adding points to your break lines IF they pas through actual points.
Further there is also no 'real' value in proximity break lines.
As the name implies, and vertex along the break line is given the level (elevation) of the nearest (proximal)
point; and in reality that is NOT the elevation of that vertex.
Be your Best


Michael Farrell
http://primeservicesglobal.com/

alanjt

  • Needs a day job
  • Posts: 5352
  • Standby for witty remark...
Re: Jagged profile from surface created from points...HELP
« Reply #5 on: September 19, 2014, 03:06:07 PM »
I suspect you were not given the entire raw data file for the lidar.

I've used some here and there and it rarely needs breaklines to 'enhance' the surface that is developed from the point cloud(s).

I see no value in adding points to your break lines IF they pas through actual points.
Further there is also no 'real' value in proximity break lines.
As the name implies, and vertex along the break line is given the level (elevation) of the nearest (proximal)
point; and in reality that is NOT the elevation of that vertex.
I was wondering about the point data along the railroad, since the only data I have there are the provided breaklines. I added the points at each breakline vertex. I know proximity are more x/y and give nothing for z, but I was more wanting to use them that way so I could add hard points to the surface, just use the breaklines to make sure it didn't triangulate improperly over the railroad, then I could simpliy/smooth the surface by having it remove points. At that point, I was grasping at straws.

I will contact the LIDAR data providers and request the missing points along the railroad. Do you think that will solve the jaggedness?
Civil 3D 2019 ~ Windohz 7 64bit
Dropbox

mjfarrell

  • Seagull
  • Posts: 14444
  • Every Student their own Lesson
Re: Jagged profile from surface created from points...HELP
« Reply #6 on: September 19, 2014, 04:32:43 PM »
yes...
I have seen various forms of attempts at 'grooming' the LIDAR data.
Most of them remove wanted as well as unwanted information.

Further, as good as LIDAR is; there are still instances wherein actual survey data is
still required to achieve proper degree of resolution for the engineering design task at hand.

In this case the LIDAR is fine to define the general terrain conditions.
However I would be reluctant to attempt to design a new rail line connecting to the existing using that data.***


*** Unless in the full dataset there is enough resolution on the rail to positively locate it, horizontally and vertically.


I would REMOVE those proximity breaklines from your surface as they are not helping and don't define actual elevations properly.

Be your Best


Michael Farrell
http://primeservicesglobal.com/

alanjt

  • Needs a day job
  • Posts: 5352
  • Standby for witty remark...
Re: Jagged profile from surface created from points...HELP
« Reply #7 on: September 19, 2014, 05:17:33 PM »
Oh, we're not changing anything with the railroad. That's one of the limits of the project area.
LIDAR has its uses, but I much prefer survey data, and in the end, we'll be using 100% survey.

Thanks for the guidance, guys.
Civil 3D 2019 ~ Windohz 7 64bit
Dropbox