Building Extraction
stu
Stu
Terje Mathisen
> I am trying to use Lastools to get some building footprints for NJ
> from the post-Sandy Flight.
>
> ftp://csc.noaa.gov/pub/DigitalCoast/lidar1_z/geoid12a/data/2478/
>
> the lidar is all class 1 (unclassified) or 2 ground
>
> Here is my process.
> las2las Convert to UTM (works fine)
> lasheight
> lasground
And why are you doing it after lasheight?
Are you using lasheight with '-replace_z', in which case lasground will
be working with a totally flat terrain?
> lasclassify
>
> The problem appears to be I am not get enough class 6's, the ones I am
> getting look good but it is only catching<5% of buildings.
lasclassify is documented to need a coarser than default area grid if
the point density is too low.
Terje
> In attached - blue are all las returns, red are building points from
> lasclassify. As you can see I need many more red points.
>
> Stu
>
>
> 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"
Jonah Sullivan
1 Classify Point Cloud
1) lassplit.exe
a) (Checked) “Merge files into one” to conduct analysis on all swaths merged into a single file in memory.
b) (selected) “by flight line” - writes data from each flightline into a separate LAS file
2) lasground.exe
a) (selected) “City or Warehouses” step size is 25 metres, defining the size of the initial search area for ground seed points
b) (selected) “Fine” setting intensifies initial ground point search for steep hills
c) (selected) “compute height” setting to write height above ground in user data for each LAS point
3) lasclassify.exe
a) (defined) Search area size = 2m – defines the size of theoretical planes that are fitted to above-ground returns.
b) (defined) Building planarity = 0.1 – requires point clusters with a theoretical plane traversing them that has a standard deviation from the plane of up to 0.1m to be classified as a building.
c) (defined) Forest ruggedness = 0.4 – requires point clusters with a theoretical plane traversing them that has a standard deviation from the plane greater than 0.4m to be classified as vegetation.
d) (defined) Ground offset = 2m – requires building point be at least 2m above ground
e) (checked) Include Gutters - Specifies whether the tool should try to detect building roofs all the way to the gutter.
f) (checked) Wide Gutters - Specifies an experimental variation for better gutter detection.
g) (Checked) no tiny buildings - Specifies whether tiny groups of planar points that may or may not correspond to buildings should be classified as buildings or not.
h) (Unchecked) no tree overlap - Specifies whether the tool should allow trees to overlap previously found roofs when classifying vegetation.
4) lasoverage.exe
a) (selected) “Flag as withheld” to designate points that are overlapping adjacent swaths using the withheld flag.
2 Building Feature Extraction
1) Lasboundary.exe
a) Filter: “drop_withheld” to exclude withheld-flagged points from analysis.
b) Filter: “keep_classification 6” to exclude all points that are not classification 6.
c) (Checked) “Merge files into one” to conduct analysis on all swaths merged into a single file in memory.
d) (Defined) concavity=1.65 – voids between points of more than 1.65m (based on nominal post-spacing of 0.66m * 2.5) will be considered to be outside of features.