Lidar to 3D-Tiles Pnts Converter

The below repository details a small set of scripts created to convert LAS lidar data to the 3d-tiles points spec:

https://github.com/mattshax/cesium_pnt_generator

The sample viewer showcases a sample las dataset with >10 million color points, and uses shell and nodejs scripts to convert and write the binary pnts data.

There are lots of room for improvements in this script - specifically the RTC coordinate conversations currently are not lining up heights correctly (a manual adjustment flag has been included but suggestions are appreciated), and the tileset.json is being generated for each chunk of data as opposed to grouping as children in a single tileset file - but perhaps a good starting point for those looking for this lidar conversion functionality.

Of course this may well change in the near future as the 3d-tile points specs develops.

Comments / improvements are of course encouraged !

Wow, that looks great! Thanks for sharing!
I tweeted out a link for people to check it out: https://twitter.com/CesiumJS/status/714467345369055232

Best,

Hannah

Hi!

I tried your scripts (with the link of the data on GitHub) for convert LAS to pnts but I have some errors at the end of the script:
ERROR: unknown argument 'stadium_all.las'
[...]
ERROR: unknown argument 'image.tif'
[...]

Do you think it's come from the version of Linux? I work on Debian GNU/Linux 7.7 (wheezy).

Emma

Hi Emma,

It could be - I only tested on Ubuntu 14.04. There may well be some differences in bash on that version.

If you install docker on your machine, you can run an Ubuntu 14.04 container and run the scripts from there:

curl -fsSL https://get.docker.com/ | sh

docker run -it ubuntu:14.04 /bin/bash

Matthew

Hi Emma,

It could be - I only tested on Ubuntu 14.04. There may well be some differences in bash on that version.

If you install docker on your machine, you can run an Ubuntu 14.04 container and run the scripts from there:

Hi

Thank you for your answer! I tried on Ubuntu 14.04 and it's work.

Emma

Glad to hear it !

BTW I just stood up a working version of the viewer at the link below - first time loading takes a bit of time:

http://points.chicago.vision/