[pdal] Trouble reading some 3DEP tiles in PDAL
Andrew Bell
andrew.bell.ia at gmail.com
Sun Sep 15 11:29:15 PDT 2024
On Thu, Sep 12, 2024 at 9:06 PM Adam Steer <adam.d.steer at gmail.com> wrote:
>
> Hi all
>
> I appreciate the approach of PDAL reporting what is wrong and saying 'nope, not proper las' - currently I'm working on a small project where that is important feedback to a vendor.
>
> The issue I can see around this style of hard rejecting a dataset is that forensic data examination is ...difficult.
In most cases the error message indicates more than "nope, not proper
LAS". Some could be more informative.
> For example 'PDAL: readers.las: Extra byte specification exceeds point length beyond base format length.' leads to lots of reading and guesswork, because I can't actually interrogate the data without knowing to use other tools.
This message could provide some numbers, but it's also pretty clear
what the problem is, IMO.
> The current workaround is LASpy -> PDAL. If it helps to know, LASpy reads in then writes the same files out as LAS which passes PDAL checks. I should get around to closing that loop today (ie reporting what is different about the LASpy-written lasfile).
We have discussed a PDAL command that would specifically validate LAS
files. Perhaps your client would be interested in funding this work?
--
Andrew Bell
andrew.bell.ia at gmail.com
More information about the pdal
mailing list