Power line and Tower Pole

[simaya35]

Dear group,

My research in Lidar height based classification for all height object (Tree, Building,Power line, Tower pole).

I have processed  ( lasheight, lasnoise,lasclassify)  lastools, but provide different results than expected. 


Trees and building point classified nice but sum more point  miss-classified power line and tower poles.

Processing las tools power line and tower pole totally miss matched, tower pole hight for 10 meter.

 How to improve this process any manual document in  particular my research related,  please share this group.

 
Already see the bellow  tutorials

http://rapidlasso.com/2013/04/20/tutorial-quality-checking/
http://rapidlasso.com/2013/10/13/tutorial-lidar-preparation/
http://rapidlasso.com/2013/10/20/tutorial-derivative-production/
http://rapidlasso.com/2014/03/02/tutorial-manual-lidar-editing/

Regards,

simaya

[Terje Mathisen]

simaya35 wrote:
> Dear group,
>
>
> My research in Lidar height based classification for all height object
> (Tree, Building,Power line, Tower pole).
>
> I have processed ( lasheight, lasnoise,lasclassify) lastools, but
> provide different results than expected.
>
>
> Trees and building point classified nice but sum more point
> miss-classified power line and tower poles.
> Processing las tools power line and tower pole totally miss matched,
> tower pole hight for 10 meter.

Your problem here is almost certainly related to the Shannon Sampling
Theorem:

Unless you get at least a couple of hits very near the top of each tower
pole, the height _must_ be misjudged!

With normal sampling rates for LiDAR (often in the 0.5 to 4 samples per
sq meter range), it is extremely easy to miss a pole completely.

What you _can_ do is to look for a faint stream of returns from the
cables between each pole:

This trace will be located along a straight line in the (x,y) plane, and
with the height distribution following the classic catenary (hyperbolic
cosine) curve, right?

You can curve fit each localized group of such air returns, estimating
the relevant constants for the curve, then you will find both the poles
and their (attachment) heights at the intersections between them.

Good luck!

Terje

>
> How to improve this process any manual document in particular my
> research related, please share this group.
>
> Already see the bellow tutorials
>
> http://rapidlasso.com/2013/04/20/tutorial-quality-checking/

> <http://rapidlasso.com/2013/04/20/tutorial-quality-checking/>
> http://rapidlasso.com/2013/10/13/tutorial-lidar-preparation/
> <http://rapidlasso.com/2013/10/13/tutorial-lidar-preparation/>
> http://rapidlasso.com/2013/10/20/tutorial-derivative-production/
> <http://rapidlasso.com/2013/10/20/tutorial-derivative-production/>
> http://rapidlasso.com/2014/03/02/tutorial-manual-lidar-editing/
> <http://rapidlasso.com/2014/03/02/tutorial-manual-lidar-editing/>
>
> Regards,
>
> simaya
> --
> Download LAStools at
> http://lastools.org
> http://rapidlasso.com
> Be social with LAStools at
> http://facebook.com/LAStools
> http://twitter.com/LAStools
> http://linkedin.com/groups/LAStools-4408378
> Manage your settings at
> http://groups.google.com/group/lastools/subscribe


--
- <Terje.M...@tmsw.no>
"almost all programming can be viewed as an exercise in caching"