[ros-users] Depth image REP

[Kevin Walchko]

-1

It is a little hard to understand this trade due to lack of info, but I will justify my answer:

- There is a major re-write of code (both core ROS and community provided code) that could take significant time and effort to perform. It would be useful if an estimate of this impact could be given (even a WAG). However, maybe I am wrong.

- There appear to be no crucial information included in the new DisparityImage to really justify the change and the effort above. At least there is no clear, compelling argument in the REP. 

- I dislike the justification being because one vendor has set a precedent. What happens when the kinect 2 comes out with a different format, will we adapt that format as the next default? What happens if OpenNI goes away and Nintendo (just picked someone at random) invents a really cool depth sensor that works different? I think our depth image should be based on what is right for us and not what one vender provides today. Just remember, the Kinect is only a little over one year old. What will next year this time bring?

I did like the idea of uint16 taking less bandwidth and potentially running on simpler hardware (ARM, netbook, etc). However, I find none of the arguments compelling.

So you are proposing:
DisparityImage of type uint16
- focal length, baseline, principal point
- valid_window (seems unnecessary)
- min/max disparity
- delta_d

I assume we would get rid of the need for CameraInfo when used with DisparityImage message if this is done.

What is the impact of instead publishing a separate topic to fill in these couple of missing metadata pieces? This changes things from (Image, CameraInfo) pair to (Image, CameraInfo, ??) triplet … not sure that is super great, but the impact may be much less. 

Also, not to be too critical, but I had a problem following the logic of this entire REP. I am sure it makes complete sense to the author, however I found the logic to jump around.

This is an interesting suggestion, lets just make sure it is the right one for ROS ... thanks!

Kevin


On Dec 6, 2011, at 5:51 PM, Patrick Mihelich wrote:

> Hi all,
> 
> Attached is a draft REP defining a representation for depth images in ROS. This standardizes the format used by the ROS OpenNI  driver and associated processing nodelets, and should be useful to other producers/consumers of depth data.
> 
> You can also view it in HTML at http://people.willowgarage.com/mihelich/rep-depth-image.html.
> 
> If this interests you, please state your vote (see REP 10 - Voting Guidelines) as well as any comments and concerns.
> 
> We seem to be having archival issues, so (light) early feedback for the ros-sig-perpip list is reproduced below.
> 
> Cheers,
> Patrick
> 
> Forwarded conversation
> Subject: Depth image REP
> ------------------------
> 
> From: Ethan Rublee <erublee at willowgarage.com>
> Date: Fri, Dec 2, 2011 at 3:56 PM
> To: Patrick Mihelich <mihelich at willowgarage.com>
> Cc: ros-sig-perpip at code.ros.org
> 
> 
> +1
> 
> ----------
> From: Vincent Rabaud <vrabaud at willowgarage.com>
> Date: Fri, Dec 2, 2011 at 4:13 PM
> To: Ethan Rublee <erublee at willowgarage.com>
> Cc: Patrick Mihelich <mihelich at willowgarage.com>, ros-sig-perpip at code.ros.org
> 
> 
> +0, it looks good but I don't know much about the previous messages. The deprecation of the DisparityImage will happen through the tic-toc I guess.
> 
> ----------
> From: Brian Gerkey <gerkey at willowgarage.com>
> Date: Fri, Dec 2, 2011 at 5:47 PM
> To: Patrick Mihelich <mihelich at willowgarage.com>
> Cc: ros-sig-perpip at code.ros.org
> 
> 
> +0.
> 
> I'm not a user of this part of the system, but find the REP to be
> well-reasoned and convincing.
> 
>        brian.
> 
> 
> <rep-depth-image.txt>_______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ros.org/pipermail/ros-users/attachments/20111206/fc081bbf/attachment-0004.html>