search

NeBula-Autonomy / nebula-odometry-dataset

Ready to test your SLAM system in challenging datasets from extreme environments? Try this out! The dataset is provided by the Team CoSTAR that has been intensively testing multi-robot systems in real world environments such as caves, tunnels, abandoned factories and industrial plants for the DARPA Subterranean Challenge.
MIT License
156 stars 15 forks source link

about timestamp #13

Open whw-create opened 2 years ago

whw-create commented 2 years ago

Didn't you record the timestamp of each point in the bag ?

BenjaminMorrell commented 2 years ago

Unfortunately not - we just have the per-scan timestamps.

There may de a few of the datasets that have the velodyne packets (which has that information), but they are not uploaded yet.

whw-create commented 2 years ago

Thank you for your reply

------------------ 原始邮件 ------------------ 发件人: "NeBula-Autonomy/nebula-odometry-dataset" @.>; 发送时间: 2022年9月27日(星期二) 凌晨4:06 @.>; @.**@.>; 主题: Re: [NeBula-Autonomy/nebula-odometry-dataset] about timestamp (Issue #13)

Unfortunately not - we just have the per-scan timestamps.

There may de a few of the datasets that have the velodyne packets (which has that information), but they are not uploaded yet.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

whw-create commented 1 year ago

hi, did you projected all lidars' point to one lidar already? cause i can't find extrinsic between lidars in your yaml of H, thanks for reply

femust commented 1 year ago

in additional config files there are all the transformations that are needed

whw-create commented 1 year ago

in additional config files there are all the transformations that are needed

i'm sorry but i really didn't find the extrinsic between lidars of dataset H in that extra files link

femust commented 1 year ago

in the file there are relative transformations between different modules (cameras/lidars/base_link ,etc), so you just need to use a chain rule since there are relative transformations (x,y,z,r,p,y (Euler)) .

Frankly speaking the best idea and the easiest way is to just read this to urdf, see how we're doing this in locus over here

whw-create commented 1 year ago

in the file there are relative transformations between different modules (cameras/lidars/base_link ,etc), so you just need to use a chain rule since there are relative transformations (x,y,z,r,p,y (Euler)) .

Frankly speaking the best idea and the easiest way is to just read this to urdf, see how we're doing this in locus over here

thanks