Re: [geogig-dev] LIDAR format support

[dp mcmlxxvi <dpmcmlxxvi@xxxxxxxxx>]

That is certainly one possible workflow where it might have limited application. The are a few others I think might benefit from something like geogig.

1. Quality Control. LIDAR gets edited during QC processes. An initial point cloud can be quite noisy, often to the point of being unusable. The point cloud is often edited either by automated algorithms (e.g., denoising), or manually by analysts, or some combination of the two. Being able to maintain that history would seem value to a QA process. For instance, reverting the results of some automated QC algorithm in order to apply another. Or merging the QC edits of different analysts.

2. Change detection. Being able to diff between two collections and create a "patch" would be another. For example, diffs with something like git can be be refined to ignore things like newlines, spaces, etc. A similar concept could be applied with something like geogig, where the diff refinement can be spatial in nature. So, two points within a certain threshold distance can be treated as equal. Therefore, the two point clouds in your example that cover the same area but the points aren't exactly the same could still be seen as "equivalent" within the diff threshold distance.

thanks

dan

On Wed, Feb 22, 2017 at 12:59 PM, Gabriel Roldan <groldan@xxxxxxxxxxxxxxxx> wrote:

Hi there,

as far as I know it's not on the roadmap.

Now, I'm far from an expert in LIDAR formats, but my impression is you wouldn't get much benefit from versioning LIDAR point clouds with geogig. Rationale being that, to the extent of my knowledge, two passes of LIDAR over the same area would produce two completely different datasets? it's not like points would lay on the exact same spot except if something changed at that very exact location. So, two point clouds would be totally different despite as a whole they'd represent roughly the same information?

In the context of vectorial datasets, which geogig is designed for, you can track change as it happens on certain features while others remain the same (e.g. editing through WFS/QGIS/etc). My impression is that model wouldn't fit for LIDAR data and every two versions would be completely different, meaning you'd have to store two completely different datasets, getting no real benefit out of geogig.

I might be wrong though, looking forward to your comments.

cheers,

Gabriel

On Mon, Feb 20, 2017 at 11:00 PM, dp mcmlxxvi <dpmcmlxxvi@xxxxxxxxx> wrote:

Greetings all,

First, great project.

Second, is supporting LIDAR formats (e.g., LAS) on the roadmap? I see importing other formats like GeoJSON is being done via GeoTools. Is that the roadblock to supporting LAS, needing a GeoTools reader/writer for LAS?

I was looking at options for geospatial versioning of LIDAR data. Has GeoGig been tested with large 3D point clouds?

Dan

https://dpmcmlxxvi.github.io

_______________________________________________
geogig-dev mailing list
geogig-dev@xxxxxxxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.locationtech.org/mailman/listinfo/geogig-dev

--

Gabriel Roldán
Software Developer | Boundless
groldan@xxxxxxxxxxxxxxxx
@boundlessgeo

_______________________________________________
geogig-dev mailing list
geogig-dev@xxxxxxxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.locationtech.org/mailman/listinfo/geogig-dev