explanation of three warnings...

[Kirschbaum, Alan]

Could anyone explain the three separate types of errors I receive while running lastile? I just want to make sure I'm not losing points or ignoring important warnings.

Thanks

-al

1) WARNING: written 2679518 points but expected 0 points

2) WARNING: on-the-fly merged files contain too many points

3) WARNING: rescaling from 0.01 to 0.001 causes LAS integer overflow for min_x

Al Kirschbaum
Remote Sensing Specialist
National Park Service

Great Lakes Inventory and Monitoring Network
2800 Lake Shore Drive East, Suite D
Ashland, WI 54806
715.682.0631 x227
al_kir...@nps.gov

[Kirk Waters - NOAA Federal]

Al,

You didn't show your original command, so I'm guessing a little on the first one.

1) Typical warning when you start from text files since there is no information about how many points to expect. Not generally a problem.

2) LAS 1.2 files can only hold 4 billion points. Your tiles are of a size that your data doesn't fit in that. Note this conflicts with my guess on #1 as it suggests you started with LAS 1.2 files. Maybe you specified them for output? Or it defaults to them?

3) Your x values are such that scaling them to integers using a factor of 0.001 exceeds the +/-2 billion limit on a 32-bit integer. Also a little silly to store millimeter precision for a laser spot that is at least a centimeter in size. Did you specify this scaling? If LAStools has a default for this, it's likely 0.01.

Kirk Waters, PhD                     | NOAA Office for Coastal Management

Applied Sciences Program      | 2234 South Hobson Ave
843-740-1227                          | Charleston, SC 29405    

coast.noaa.gov/digitalcoast

--
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

[Kirschbaum, Alan]

Sorry for not including the original command.....

lastile -i -v f:\AshlandCo_WI_LIDAR_Ayres2\Classified_LAS\*.las ^

    -drop_scan_angle_above 20^

    -drop_scan_angle_below -20^

    -tile_size 2000 -buffer 300 ^

    -odir %BLOCK%\laz_tiled -olaz ^

    -cores %CORES%

Al Kirschbaum
Remote Sensing Specialist
National Park Service

Great Lakes Inventory and Monitoring Network
2800 Lake Shore Drive East, Suite D
Ashland, WI 54806
715.682.0631 x227
al_kir...@nps.gov

[Kirschbaum, Alan]

​And, here are a couple lasinfo outputs. 

original_pretile_info.txt: an example of one of the original input las tiles that will be tiled

lastile_output​_tile_info.txt: an example of a tile resulting from lastile

Al Kirschbaum
Remote Sensing Specialist
National Park Service

Great Lakes Inventory and Monitoring Network
2800 Lake Shore Drive East, Suite D
Ashland, WI 54806
715.682.0631 x227
al_kir...@nps.gov

[Martin Isenburg]

Hello,

so it seems like you have a whole bunch of input files (such as the one in the "original_pretile_info.txt") that sum up to over 10 billion points and (at least some of them) have scale factors of 0.001 (millilmeter of millifeet) for the coordinates. You get this

2) WARNING: on-the-fly merged files contain too many points

because LAStools although can on-the-fly '-merge' 10 billion points - if you were to store the resulting output to a single LAS 1.2 file - like by using las2las - you'd be in trouble.  You get this

1) WARNING: written 2679518 points but expected 0 points

because the tile writer has no clue how many points will go into each tile. So he writes 0 in the header and when the writing of points is done the mismatch (between 0 and the actual number) is automatically reported by the library. I should maybe remove the warning in this (literally guaranteed) case. You get this

3) WARNING: rescaling from 0.01 to 0.001 causes LAS integer overflow for min_x

because you probably have a mix of resolutions in your input files. By default the resolution of the first file in the *.laz list is used for the on-the-fly '-merge' and that seems to be 0.001 so that later files that have 0.01 resolution need to have their internally stored integers multiplied with 10 which can potentially create an integer overflow.

Look in the lasinfo output in "original_pretile_info.txt" where it reports

WARNING: there is coordinate resolution fluff (x10) in XYZ

That means that although the scale factor 0.001 promises millimeter (or millifoot) resolution the coordinates are actually all multiples of 10 so that the last decimal digit of the millimeter (or millifeet) number is always zero. Hence there really should be a scale factor of 0.01 as there is only centimeter (or centifoot) resolution in the file anyways. Bad bad 'GeoCue GeoCoder' ... (-;

You can fix that by simply enforcing centimeter (or centifoot) resolution. Something I recommend doing for all airborne point clouds that have scanning errors of plus / minus a few centimeters. Here is your new command-line

lasindex -i f:\AshlandCo_WI_LIDAR_Ayres2\Classified_LAS\*.las ^

              -cores %CORES%

lastile -i -v f:\AshlandCo_WI_LIDAR_Ayres2\Classified_LAS\*.las ^

    -rescale 0.01 0.01 0.01 ^

    -drop_scan_angle_above 20 ^

    -drop_scan_angle_below -20 ^

    -tile_size 2000 -buffer 300 -flag_as_withheld ^

    -odir %BLOCK%\laz_tiled -olaz ^

    -cores %CORES%

The other additions '-flag_as_withheld' makes it easy to '-drop_withheld' later to get rid of the buffers.

Regards,

Martin

PS: A buffer of 300 seems a little excessive. Large water bodies?

[Kirschbaum, Alan]

Many thanks for the explanations Martin. I also appreciate the suggestions. The 300 buffer is in feet...do you have a general rule of thumb to use for the buffer size? For example should it be 10% of the tile size? 

-al

Al Kirschbaum
Remote Sensing Specialist
National Park Service

Great Lakes Inventory and Monitoring Network
2800 Lake Shore Drive East, Suite D
Ashland, WI 54806
715.682.0631 x227
al_kir...@nps.gov

[Martin Isenburg]

Hello,

probably the most important one is lasground. In case you run lasground or lasground_new then the buffer size must be bigger than the chose step size. You see the step size reported in the when you run in verbose '-v' mode. For lasground the step size is 1 meter for '-archeology' mode, 1 meter for '-wilderness' mode, 5 meter for '-nature' mode, 10 meter for '-town' mode, 25 meter for '-city' mode and 50 meter for '-metro' mode. If you run las2dem and you have large water bodies then you need a buffer large enough to bridge across the water to the opposite shore if you want to close the (possibly empty) water areas. Similar with any other empty return area (IKEAs and other large ware houses when creating a DTM).

Regards.

Martin @rapidlasso