-
Notifications
You must be signed in to change notification settings - Fork 185
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add docs for Denoise Layer plugin (costmap_2d) #209
Add docs for Denoise Layer plugin (costmap_2d) #209
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is pretty large plugin being adding into Nav2 stack. So, please add new "New Plugins" section to "Galactic to Humble" migration guide where to make a short description and reference to the Denoise Layer plugin there. You can find similar sections in Galactic and Foxy migration guides as an example.
Also, there are some minor fixes from below before one can go into mainstream.
I have fixed the issues mentioned. If at the same time I did not introduce new shortcomings, then I am ready to synchronize (fetch + rebase) whis the parent branch. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you for the thorough work there! Some small changes are required and we could go.
59b9f7f
to
6513710
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
1. Enable Denoise Layer | ||
----------------------- | ||
|
||
Denoise Layer is Costmap2D plugin. You can enable the ``DenoiseLayer`` plugin in Costmap2D by adding ``denoise_layer`` to the ``plugins`` parameter in ``nav2_params.yaml``. You can place it in the ``global_costmap`` and (or) ``local_costmap`` to filter noise on a global or local map. The DenoiseLayer plugin should have the following parameter defined: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I would add a note here that this is typically most usable in the local costmap, though usable in the global costmap as well. Since the local costmap updates are much smaller, it will have less impact on run-time performance. I'm concerned that if users generally use this in the global costmap, it will cause performance issues when paired with typically-high-range lidars of 20+ meters which the update window may be massive.
Another reason is that since this works only in the update window, if you have salt and pepper noise in the local costmap, it disappears and is not remembered when going back into that region since the local costmap doesn't store data in areas after it has left.
The global costmap does though, so if we leave an area and the salt / pepper noise was being handled by this later, it would still be there when we came back. This isn't necessarily a problem, but makes me want to make the point to users that you can do this, but there are some real things to think about if doing so.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done. If I misunderstood your thought or described it poorly, please help me with a clear statement of this warning.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done in 68827cc after rebase
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please re-base over latest main branch; and check Steve and my comments
6513710
to
279ed83
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The documentation updates seems to be already ironed in previous reviews. There are a few strokes and I see no problem to go.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
One comment (please restore back clarification). After it, LGTM.
@SteveMacenski, could you also please take a look on the PR?
4ac0320
to
4ca7a45
Compare
@SteveMacenski, @AlexeyMerzlyakov, does this pull request still have a chance of being accepted? If yes, I will keep this PR relevant for the main branch. |
Yes, Alexey messaged me last week about reviewing this after he's given his approvals and its my plan to do it the week after next (unfortunately, I'm no longer at Samsung Research so my time for reviewing large scale changes is limited between other responsibilities, but I'm committed to having your work represented in Nav2 given how much time you've put into it). @AlexeyMerzlyakov what would help me greatly is if you reviewed your review comments and "Resolve" the ones that are completed to your satisfaction already so that they're no longer expanded in the thread. That way, I can look at only the ones that are not yet complete while I read through the full thread in detail as I'm reviewing the code for context. It would be good to know with certainty which are "done" and which might have been forgotten about. Ditto for the nav2 repo PR. |
It looks like I have no access or rights to do that in this repository. The PR review items do not have "Resolve" button anymore: |
Ask and ye shall be given (you should receive an email for getting write access). You should see them once you accept |
@SteveMacenski, thank you for the support! |
Remove those 3 lines (and maybe reword the "Firstly" since now there's only 1 item?) and this LGTM after fixing the merge conflict Great job on the tutorial, its rare I read one where I have no substantive remarks. I had some wording nitpicks but I took care of those myself, at least those that I could from the github UI. |
Signed-off-by: ryzhikovas <ryzhikovas@gmail.com>
Signed-off-by: ryzhikovas <ryzhikovas@gmail.com>
7175bd7
to
3cd2571
Compare
Thank you for the exact wording. The small number of edits is the merit of Alexey, who helped a lot with the documentation at an early stage. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Approved pending code review itself
Docs for this feature.