/kitti2bag

Convert KITTI dataset to ROS bag file the easy way!

Primary LanguagePythonMIT LicenseMIT

kitti2bag

Hello everybody! I'm looking for more people that can bring this package to the next level. If you'd like to help you can contact me via the email. I'll be happy for every news contribution!

Gitter Build Status

Convert KITTI dataset to ROS bag file the easy way!

KITTI playback preview

Collaboration

This package enjoyed significant interest from more people then I could have thought at the beginning. I'm really glad to see that. I see many PRs and issues being raised but my day job does not allow me to push this repository further. In order to allow this package to prosper, I'm opening it up for the community. I'm more than happy to add you as a collaborator to this repository. Just send me an email.

And by the way. To ensure we maintain the quality of the repo you are required to get the PR approval from at least one other collaborator. You can use the Gitter to communicate with others.

TODOs

Help me make this feature rich and complete. Just fork this repo, implement new features (very easy in this case) and make pull request.

Feature request list:

  • make URDF of a car so transformations between frames are easily done by ROS itself.
  • deal with tracklets
  • support for unsynced+unrectified version
  • provide documentation via ROS wiki
  • provide simple GUI
  • distribute publically available bagfiles (is there a reliable public storage for this purpose?)
  • export only subset of sensors

Contributions

Thanks to the work of @jnitsch, kitti2bag can now export velodyne laser data and dynamic tf transformations. Thanks to @emreay-, this tool can now convert odometry datasets too. Thank you both!

How to install it?

It is very easy! On the machine with ROS installed, just run

pip install kitti2bag

How to run it?

One example is better then thousand words so here it is

$ wget https://s3.eu-central-1.amazonaws.com/avg-kitti/raw_data/2011_09_26_drive_0002/2011_09_26_drive_0002_sync.zip
$ wget https://s3.eu-central-1.amazonaws.com/avg-kitti/raw_data/2011_09_26_calib.zip
$ unzip 2011_09_26_drive_0002_sync.zip
$ unzip 2011_09_26_calib.zip
$ kitti2bag -t 2011_09_26 -r 0002 raw_synced .
Exporting static transformations
Exporting time dependent transformations
Exporting IMU
Exporting camera 0
100% (77 of 77) |##########################| Elapsed Time: 0:00:00 Time: 0:00:00
Exporting camera 1
100% (77 of 77) |##########################| Elapsed Time: 0:00:00 Time: 0:00:00
Exporting camera 2
100% (77 of 77) |##########################| Elapsed Time: 0:00:01 Time: 0:00:01
Exporting camera 3
100% (77 of 77) |##########################| Elapsed Time: 0:00:01 Time: 0:00:01
Exporting velodyne data
100% (77 of 77) |##########################| Elapsed Time: 0:00:15 Time: 0:00:15
## OVERVIEW ##
path:        kitti_2011_09_26_drive_0002_synced.bag
version:     2.0
duration:    7.8s
start:       Sep 26 2011 13:02:44.33 (1317042164.33)
end:         Sep 26 2011 13:02:52.16 (1317042172.16)
size:        417.2 MB
messages:    1078
compression: none [308/308 chunks]
types:       geometry_msgs/TwistStamped [98d34b0043a2093cf9d9345ab6eef12e]
             sensor_msgs/CameraInfo     [c9a58c1b0b154e0e6da7578cb991d214]
             sensor_msgs/Image          [060021388200f6f0f447d0fcd9c64743]
             sensor_msgs/Imu            [6a62c6daae103f4ff57a132d6f95cec2]
             sensor_msgs/NavSatFix      [2d3a8cd499b9b4a0249fb98fd05cfa48]
             sensor_msgs/PointCloud2    [1158d486dd51d683ce2f1be655c3c181]
             tf2_msgs/TFMessage         [94810edda583a504dfda3829e70d7eec]
topics:      /kitti/camera_color_left/camera_info    77 msgs    : sensor_msgs/CameraInfo    
             /kitti/camera_color_left/image_raw      77 msgs    : sensor_msgs/Image         
             /kitti/camera_color_right/camera_info   77 msgs    : sensor_msgs/CameraInfo    
             /kitti/camera_color_right/image_raw     77 msgs    : sensor_msgs/Image         
             /kitti/camera_gray_left/camera_info     77 msgs    : sensor_msgs/CameraInfo    
             /kitti/camera_gray_left/image_raw       77 msgs    : sensor_msgs/Image         
             /kitti/camera_gray_right/camera_info    77 msgs    : sensor_msgs/CameraInfo    
             /kitti/camera_gray_right/image_raw      77 msgs    : sensor_msgs/Image         
             /kitti/oxts/gps/fix                     77 msgs    : sensor_msgs/NavSatFix     
             /kitti/oxts/gps/vel                     77 msgs    : geometry_msgs/TwistStamped
             /kitti/oxts/imu                         77 msgs    : sensor_msgs/Imu           
             /kitti/velo/pointcloud                  77 msgs    : sensor_msgs/PointCloud2   
             /tf                                     77 msgs    : tf2_msgs/TFMessage        
             /tf_static                              77 msgs    : tf2_msgs/TFMessage

That's it. You have file kitti_2011_09_26_drive_0002_sync.bag that contains your data.

Other source files can be found at KITTI raw data page.

If you got an error saying something like command not found it means that your python installation is in bad shape. You might try running python -m kitti2bag -t 2011_09_26 -r 0002 raw_synced . Or maybe use Docker.

Prefer Docker?

That is easy too. There is a pre-built image tomas789/kitti2bag.

$ wget https://s3.eu-central-1.amazonaws.com/avg-kitti/raw_data/2011_09_26_drive_0002/2011_09_26_drive_0002_sync.zip
$ wget https://s3.eu-central-1.amazonaws.com/avg-kitti/raw_data/2011_09_26_calib.zip
$ unzip 2011_09_26_drive_0002_sync.zip
$ unzip 2011_09_26_calib.zip
$ docker run -v `pwd`:/data -it tomas789/kitti2bag -t 2011_09_26 -r 0002 raw_synced
Exporting static transformations
Exporting time dependent transformations
...

This might also be a better alternative if you are having troubles installing the package.

Bug reporting, support and feature requests.

I appreciate pull requests with bug fixes and new features. You you want to help with something please use GitHub issue tracker.

Related works

  • pykitti is very simple library for dealing with KITTI dataset in python.
  • kitti_player allows to play dataset directly. No bag file needed. I found difficult to get it work. Some bug fixed can be found in my fork of kitti_player but still not good enough.