[Liblas-devel] Re: [GRASS-user] LiDAR LAS import

[Howard Butler]

Awesome news!  I'm glad the GRASS community (er, Markus rather) came together to get something done.  I hope it serves its purpose well and beats las2txt | v.in.ascii by a lot.

A few points I would like to clear up:

- libLAS only supports 1.0, 1.1, 1.2, and 1.3 with support for Point Formats 0-3 in those versions at this time
- libLAS has not been widely tested on big endian platforms and it is expected to work, but might not :)
- OGR will most likely never support LAS or LAZ data natively due to the fact that it isn't practical to stuff millions of individual points into OGR's data model.
- I hope that Markus' v.in.lidar ingester is not doing Seek() as part of its reading strategy.  Seek() on LAZ data is going to be horribly slow (much faster in an upcoming version though!).

Congrats Markus!

Howard

On May 25, 2011, at 5:17 AM, Hamish wrote:

> [fyi, --H]
> 
> 
> Markus Metz wrote:
>> Hi all,
>> 
>> GRASS 7 has a new module v.in.lidar for importing LiDAR LAS files
>> (*.las or *.laz). The LAS file format is commonly used for storing
>> LiDAR point clouds, but is unfortunately not supported by OGR.
>> v.in.lidar uses the libLAS library [0] and is only compiled if the
>> libLAS library is present.
>> 
>> I chose to use the library instead of writing a custom LAS reading
>> interface because the current LAS library version 1.6.1 is stable,
>> supports LAS file versions 1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, each of
>> which can store LiDAR points in up to 5 different point formats. The
>> user and the interface do not need to know the file version and point
>> format of a given file, all that is conveniently handled by the libLAS
>> library in the background. The library has Large File Support (LFS)
>> and is well tested on different platforms, also with different
>> endian-ness. This functionality is not that easy to replicate.
>> 
>> You will need to get the libLAS library and configure GRASS 7 with
>> --with-liblas in order to have the module available. Please test!
>> 
>> Markus M
>> 
>> [0] http://www.liblas.org
> 
> 
> neat! any time trials to say how much faster it is than piping
> las2txt | v.in.ascii
> ?
> 
> 
> fyi I have plans (but more ideas than time I'm afraid) to also add a
> stream= option to r.in.xyz with the possible options of ascii, las, mbio.
> There'll probably be another option added to select signal strength, etc.
> but I might overload the z= column parameter to accept strings for that(?).
> 
> (MBIO is for MB-System's multibeam sonar input library, Bob Covill already
> has this integrated in a custom fork)
> 
> 
> cheers,
> Hamish
> 
> 
>> _______________________________________________
>> grass-user mailing list
>> grass-user at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/grass-user
> _______________________________________________
> Liblas-devel mailing list
> Liblas-devel at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/liblas-devel