search

NeoGeographyToolkit / StereoPipeline

The NASA Ames Stereo Pipeline is a suite of automated geodesy & stereogrammetry tools designed for processing planetary imagery captured from orbiting and landed robotic explorers on other planets.
Apache License 2.0
479 stars 168 forks source link

Standalone point cloud erode? #288

Closed foobarbecue closed 3 years ago

foobarbecue commented 4 years ago

It's nice that dem_mosaic and point2dem have "erode" options which remove points from the edges of clouds.

I would like to do that to my point clouds before aligning several clouds using pc_align or n_align.

Is it possible to just run erode?

oleg-alexandrov commented 4 years ago

No, we provide no facilities to erode points clouds by themselves. You could try to convert them to las with point2las and look for any tools that process this format (there should be some, as it is a pretty widespread format). Then point2dem or pc_align could use your cleaned up las file.

On Tue, Nov 12, 2019 at 3:09 PM Aaron Curtis notifications@github.com wrote:

It's nice that dem_mosaic and point2dem have "erode" options which remove points from the edges of clouds.

I would like to do that to my point clouds before aligning several clouds using pc_align or n_align.

Is it possible to just run erode?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/NeoGeographyToolkit/StereoPipeline/issues/288?email_source=notifications&email_token=AAKDU3EUFBEV2AMTV7HWM2TQTMZTLA5CNFSM4JMKVHU2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HY24LWQ, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAKDU3AXOSQ2PJQUVUSXEUDQTMZTLANCNFSM4JMKVHUQ .

foobarbecue commented 4 years ago

I think the reason I felt I needed this was possibly because of spurious points caused by what's happening in #302

oleg-alexandrov commented 4 years ago

Just for the record putting here a small comment I put at that issue as well. You can try to nip this in the bud, so to speak, by doing erosion at stereo_fltr stage, with --erode-max-size. That one at some point we should make positive as the default.

oleg-alexandrov commented 4 years ago

(And this does not mean all stereo must be redone, just the fltr step and what comes next, so one should use stereo -e 3 to resume at this step.)

oleg-alexandrov commented 3 years ago

I added to point2las the ability to filter out points by triangulation error, as for point2dem. That should help a bit with filtering the cloud, and then the las file can be passed to pc_align. Note quite what asked, but a step on the way.