Hello,

the links for downloading the dataset in the "How to run it?" section are not working:

• wget http://kitti.is.tue.mpg.de/kitti/raw_data/2011_09_26_drive_0002/2011_09_26_drive_0002_sync.zip

• wget http://kitti.is.tue.mpg.de/kitti/raw_data/2011_09_26_calib.zip

Looking at the links at the KIT website it seems they moved their files (https://s3.eu-central-1.amazonaws.com/avg-kitti/raw_data/2011_09_26_drive_0001/2011_09_26_drive_0001_sync.zip)

Best,
Alex

kitti2bag is successfully installed after using pip(9.0.1), but when I execute 'kitti2bag -t 2011_09_26 -r 0001 raw_synced .'
I got mistakes as belows:

Traceback (most recent call last):
  File "/usr/local/bin/kitti2bag", line 9, in <module>
    load_entry_point('kitti2bag==1.5', 'console_scripts', 'kitti2bag')()
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 542, in load_entry_point
    return get_distribution(dist).load_entry_point(group, name)
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 2569, in load_entry_point
    return ep.load()
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 2229, in load
    return self.resolve()
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 2235, in resolve
    module = __import__(self.module_name, fromlist=['__name__'], level=0)
ImportError: No module named kitti2bag.__main__

when I run rostopic echo /kitti/velo/pointcloud/header
the seq always equals to 0, do you now how to solve it to make the seq increase frame by frame.

---
seq: 0
stamp: 
  secs: 1317671741
  nsecs: 258542060
frame_id: "velo_link"
---
seq: 0
stamp: 
  secs: 1317671741
  nsecs: 362260103
frame_id: "velo_link"
---
seq: 0
stamp: 
  secs: 1317671741
  nsecs: 465951919
frame_id: "velo_link"

Hi,

Can you please tell me how you managed the timestamps of each sensors ?

Thanks

Hi,

I wanted to convert the odometry benchmark data to bags, but since it's currently not possible to convert the point clouds from that set I used the already proposed solution, namely to convert the corresponding raw data sequences. However I also need the ground truth data for comparison. Is there a way of converting just the ground truth data, for example using the -s flag?

Hi, when I converted the kitti data to bag file using kitti2bag, I met the following error.

Traceback (most recent call last):
  File "/usr/local/bin/kitti2bag", line 16, in <module>
    import rosbag
  File "/opt/ros/melodic/lib/python2.7/site-packages/rosbag/__init__.py", line 33, in <module>
    from .bag import Bag, Compression, ROSBagException, ROSBagFormatException, ROSBagUnindexedException
  File "/opt/ros/melodic/lib/python2.7/site-packages/rosbag/bag.py", line 55, in <module>
    import gnupg
  File "/usr/local/lib/python2.7/site-packages/gnupg/__init__.py", line 23, in <module>
    from . import gnupg
  File "/usr/local/lib/python2.7/site-packages/gnupg/gnupg.py", line 42, in <module>
    from ._meta    import GPGBase
  File "/usr/local/lib/python2.7/site-packages/gnupg/_meta.py", line 134, in <module>
    class GPGBase(object):
  File "/usr/local/lib/python2.7/site-packages/gnupg/_meta.py", line 72, in __new__
    if cls._find_agent():
  File "/usr/local/lib/python2.7/site-packages/gnupg/_meta.py", line 115, in _find_agent
    if (proc.name() == "gpg-agent") and proc.is_running():
  File "/usr/local/lib/python2.7/site-packages/psutil/__init__.py", line 610, in name
    cmdline = self.cmdline()
  File "/usr/local/lib/python2.7/site-packages/psutil/__init__.py", line 663, in cmdline
    return self._proc.cmdline()
  File "/usr/local/lib/python2.7/site-packages/psutil/_psosx.py", line 342, in wrapper
    return fun(self, *args, **kwargs)
  File "/usr/local/lib/python2.7/site-packages/psutil/_psosx.py", line 426, in cmdline
    return cext.proc_cmdline(self.pid)
  File "/usr/local/Cellar/python@2/2.7.15_1/Frameworks/Python.framework/Versions/2.7/lib/python2.7/contextlib.py", line 35, in __exit__
    self.gen.throw(type, value, traceback)
  File "/usr/local/lib/python2.7/site-packages/psutil/_psosx.py", line 373, in catch_zombie
    raise ZombieProcess(proc.pid, proc._name, proc._ppid)
psutil._exceptions.ZombieProcess: psutil.ZombieProcess process still exists but it's a zombie (pid=62522)

Hi there,
I'm very new to this,
I tried installing and running your package and it works fine.
However when I try to change a topic name in kiiti2bag.py lines 284 to 287, nothing changes in the output rosbag, please advise?
Thanks a lot

Does anybody know where the world reference system of /tf is centered?
I am trying to convert /kitti/oxts/gps/fix (NavSatFix - lat, long, alt) in the /tf world (x, y, z) reference system.

Thanks.

I see there are some LIO tested on kitti, but I find there are no oxt relevant data in the odometry raw data.

@tomas789 Dear sir, when I successfully finished all the previous installation procedures and run the command "kitti2bag -t 2011_09_26 -r 0002 raw_synced .". There is an error "kitti2bag: command not found". I checked the kitti2bag package with "pip list", which was installed successfully.

Hi Tomas,
Firstly, thanks for developing this amazing package.
Regarding my issue, when I use velodyne cloud topic in a SLAM I am testing the following jumps in tf occur:

Could it be a issue of the bag conversion?
Thanks,
Bruno

Hi,

I am working with Kitti dataset and I want to add a "ring id" label to each point . Instead of having XYZI points, I am wondering if we can have XYZRI points ?

Would it be possible to just transform odometry dataset (velodyne laser data, 80 GB) into .bag file? There are 3 parameters to select: {raw_synced,odom_color,odom_gray} , but I have no idea which one to choose if I just want to save laser point into .bag.
Any suggestions?

Hi Tomas,

Thanks for your work. When I checked the code, there was node option to support the Velodyne laser data, I wonder if it is supported?

Hi,I followed the Readme introduction to convert Kitti datasets,but when I execute 'kitti2bag -t 2011_09_30 -r 0027 raw_synced .',I got mistakes as belows:

Traceback (most recent call last):
  File "/usr/local/bin/kitti2bag", line 394, in <module>
    main()
  File "/usr/local/bin/kitti2bag", line 302, in main
    kitti = pykitti.raw(args.dir, args.date, args.drive)
  File "/usr/local/lib/python2.7/dist-packages/pykitti/raw.py", line 34, in __init__
    self._load_calib()
  File "/usr/local/lib/python2.7/dist-packages/pykitti/raw.py", line 232, in _load_calib
    data['T_velo_imu'] = self._load_calib_rigid('calib_imu_to_velo.txt')
  File "/usr/local/lib/python2.7/dist-packages/pykitti/raw.py", line 147, in _load_calib_rigid
    data = utils.read_calib_file(filepath)
  File "/usr/local/lib/python2.7/dist-packages/pykitti/utils.py", line 74, in read_calib_file
    key, value = line.split(':', 1)
ValueError: need more than 1 value to unpack

Please tell me how to fix it,thank you!

What a bad program

Hello

I'm not sure what does /tf message contains. Does it containt the pose of the left camera obtained with GPS/IMU fusion?

Can I use it as ground truth?

I tried to install kitti2bag from the source repository, using the setup.py script. While running "python setup.py install" seemed to work, running kitti2bag after that results in the following error:

Traceback (most recent call last):
File "/home/thully/anaconda3/bin/kitti2bag", line 11, in
load_entry_point('kitti2bag==1.5', 'console_scripts', 'kitti2bag')()
File "/home/thully/anaconda3/lib/python3.7/site-packages/pkg_resources/init.py", line 489, in load_entry_point
return get_distribution(dist).load_entry_point(group, name)
File "/home/thully/anaconda3/lib/python3.7/site-packages/pkg_resources/init.py", line 2793, in load_entry_point
return ep.load()
File "/home/thully/anaconda3/lib/python3.7/site-packages/pkg_resources/init.py", line 2411, in load
return self.resolve()
File "/home/thully/anaconda3/lib/python3.7/site-packages/pkg_resources/init.py", line 2417, in resolve
module = import(self.module_name, fromlist=['name'], level=0)
ModuleNotFoundError: No module named 'kitti2bag'

I tried this with Python 2 as well as on a system without Anaconda installed, and get the same error. Installing using pip works, though isn't desired as I wanted to work on some potential enhancements to the script (in particular, handling tracklets)

When I tried to do the example, it is said that Could not load module pykitti. Please run pip install pykitti, then I run pip install pykitti , it is said that Requirement already satisfied.

Any suggestions?
`

When I pip install kitti2bag, got the problem:

ImportError("Tornado requires an up-to-date SSL module. This means "
ImportError: Tornado requires an up-to-date SSL module. This means Python 2.7.9+ or 3.4+ (although some distributions have backported the necessary changes to older versions).

I solved it by following this command:

sudo pip install 'Tornado>=4.0.0,<5.0.0'

Reference:

JuliaDocs/Documenter.jl#659 (comment)

you throw an error that 'please run pip install pykitti' when a importError happens
but that is not absolutely meaning we have not installed pykitti
it might be caused by a low version of numpy

input

$ python

import pykitti

to see what is the real reason

I have observed that the output sensor_msgs/CameraInfo messages have their height and width fields flipped, while the sensor_msgs/Image messages do not.

I am running OpenCV 3 and with the latest version of kitti2bag, and have not tested if this issue occurs with OpenCV 2.x.

Hello,
I wanted to try the example. But when I enter the line
wget http://kitti.is.tue.mpg.de/kitti/raw_data/2011_09_26_drive_0002/2011_09_26_drive_0002_sync.zip
I get the following error:
--2018-09-06 16:29:10-- http://kitti.is.tue.mpg.de/kitti/raw_data/2011_09_26_drive_0002/2011_09_26_drive_0002_sync.zip Auflösen des Hostnamen »kitti.is.tue.mpg.de (kitti.is.tue.mpg.de)«... fehlgeschlagen: Der Name oder der Dienst ist nicht bekannt. wget: kann die Host-Adresse »kitti.is.tue.mpg.de« nicht auflösen

I think the website is not available. Is there another way to get the sample data?

Hi,

I'm unable to generate a rosbag from the KITTI odometry Velodyne dataset (http://www.cvlibs.net/download.php?file=data_odometry_velodyne.zip).

I get a AttributeError: odometry instance has no attribute 'load_calib' when I try to run kitti2bag odom_color -s 01.

Has anyone been able to do this successfully?

Thanks!

Hi,in Readme introduction,it said "Thanks to @emreay-, this tool can now convert odometry datasets too." but What are the specific instructions for executing this method？Could anyone help me?Thank you.

How can I only convert velodyne data (*.bin) to rosbag? Only need to convert velodyne data, without image data.
I can use it to convert raw_data to rosbag.
But I can not use it to convert only odometry data in this website odometry:
In other words, this tools only can convert data that meets the formatting requirements( format in raw_data .
How can I convert odometry sequences in this website to rosbag?

So how to use folder "velodyne_points" without calibration file or image folder?
Looking forward to any reply

Just followed instruction in README and got the following error:

docker run -v `pwd`:/data -it tomas789/kitti2bag -t 2011_09_26 -r 0002 raw_synced 
Traceback (most recent call last):
  File "/usr/local/bin/kitti2bag", line 5, in <module>
    from pkg_resources import load_entry_point
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 2927, in <module>
    @_call_aside
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 2913, in _call_aside
    f(*args, **kwargs)
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 2940, in _initialize_master_working_set
    working_set = WorkingSet._build_master()
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 637, in _build_master
    return cls._build_from_requirements(__requires__)
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 650, in _build_from_requirements
    dists = ws.resolve(reqs, Environment())
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 834, in resolve
    raise VersionConflict(dist, req).with_context(dependent_req)
pkg_resources.ContextualVersionConflict: (numpy 1.11.0 (/usr/lib/python2.7/dist-packages), Requirement.parse('numpy>=1.12.0'), set(['pandas']))

I don't understand the cause of the error exactly but pandas dropped support for python2 in version 0.24 (so much for semver), and pykitti specifies pandas as a dependency without a version (so much for semver. Semver is dead, long live semver.)

Thanks for sharing codes.
I pulled the docker image of tomas789/kitti2bag then run docker, but the following error occurred.
So, tomas789/kitti2bag needs to be updated.

$ docker run -v `pwd`:/data -it tomas789/kitti2bag -t 2011_09_26 -r 0002 raw_synced
Traceback (most recent call last):
  File "/usr/local/bin/kitti2bag", line 5, in <module>
    from pkg_resources import load_entry_point
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 2927, in <module>
    @_call_aside
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 2913, in _call_aside
    f(*args, **kwargs)
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 2940, in _initialize_master_working_set
    working_set = WorkingSet._build_master()
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 637, in _build_master
    return cls._build_from_requirements(__requires__)
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 650, in _build_from_requirements
    dists = ws.resolve(reqs, Environment())
  File "/usr/lib/python2.7/dist-packages/pkg_resources/__init__.py", line 834, in resolve
    raise VersionConflict(dist, req).with_context(dependent_req)
pkg_resources.ContextualVersionConflict: (numpy 1.11.0 (/usr/lib/python2.7/dist-packages), Requirement.parse('numpy>=1.12.0'), set(['pandas']))

It seems the OXTS poses are being specified as the base_link rather than their dedicated frame imu_link.

The KITTI page shows that the GPS and IMU are in the same location.

I believe this may be the problem behind

• #8
• #11
• #12 (comment)

Please let me know if this is correct!

In order to calculate the scanline Id one has to use the "algorithm":

1. omega = asin(Z/norm(P))
2. Lookup omega in the UserManual which gives the angle increments for each laser sensor

However when I apply step 1 to the points generated from this python project I get omega values which are outside of the manufacturer supplied threshold of +2 to -24.8 degrees.

This leads me to believe there is a bug when reading the binary velodyne data.

Can anybody confirm or deny this?

Best,
Marc

This seems can only run in linux.
But how to run it in windows?

hi, I want to compare the trajectory of my algorithm with the ground truth in the kitti odometry data set, i need to project my trajectory into the coordinate system of the left camera after rectification. To verify that the projection is correct or not,replace my trajectory with the output of the GPS/IMU poses in the kitti raw data set,if the conversion is correct, the output should be consistent with the ground truth in the kitti odometry data set. Formula is as follows：
T_cam0_to_world = kitti.oxts.T_w_imu.dot(inv(kitti.calib.T_cam0_imu))
But the result is not ideal. Can you point out my mistakes and correct them?

Hello, maintainers. @tomas789 @emreay- @jnitsch .
I'm interested in contribution to develop "deal with tracklets" as a feature request.
I'm going to make a Pull-Request for that.
So, I check the Kitti Tracklet file and I think it is good to use this util from here

We can use the module as parser like following codes.

from parseTrackletXML import Tracklet,parseXML

tracklets = parseXML('data/2011_09_26/2011_09_26_drive_0002_sync/tracklet_labels.xml')

After this parsing progress, what kind of the type we need for kitti2bag ?
How about just making a custom topic message for item of tracklet_labels.xml?

If you have a good other way, please recommend it.

I'm running this package on Ubuntu 16.04 with python aliased to python2.7, I ran into this exception message:
Traceback (most recent call last): File "/usr/local/bin/kitti2bag", line 394, in <module> main() File "/usr/local/bin/kitti2bag", line 359, in main kitti = pykitti.odometry(args.dir, args.sequence) File "/usr/local/lib/python2.7/dist-packages/pykitti/odometry.py", line 34, in __init__ self._load_calib() File "/usr/local/lib/python2.7/dist-packages/pykitti/odometry.py", line 172, in _load_calib data['T_cam0_velo'] = np.reshape(filedata['Tr'], (3, 4)) KeyError: 'Tr'
How to resolve this question

There is few things that are not clear to me:

• Where can I find sections navstat_to_string, gps_mode_to_string, gps_mode_to_string, gps_mode_to_string?
• How to determine if it is GPS fix see status
• How to determine position covariance matrix correctly see position_covariance?
• Is it ok to provide pos_accuracy^2 as diagonal part of position_covariance field and set position_covariance_type to COVARIANCE_TYPE_DIAGONAL_KNOWN?

Hi.
I'm trying to use your converter.
Ubuntu 16.04.
Python 2.7.12 & 3.5.2
ROS installed

kitti2bag was installed using pip

Executing next string
kitti2bag -t 2011_09_26 -r 0001 raw_synced .

always leads to

Segmentation fault (core dumped)

Hi, I pay attention to that in kitti2bag.py, imu message is assigned as

        imu.linear_acceleration.x = oxts.packet.af
        imu.linear_acceleration.y = oxts.packet.al
        imu.linear_acceleration.z = oxts.packet.au
        imu.angular_velocity.x = oxts.packet.wf
        imu.angular_velocity.y = oxts.packet.wl
        imu.angular_velocity.z = oxts.packet.wu

But in the definition of kitti raw_data:

af: forward acceleration (m/s^2).
al: leftward acceleration (m/s^2).
au: upward acceleration (m/s^2).
wf: angular rate around forward axis (rad/s).
wl: angular rate around leftward axis (rad/s).
wu: angular rate around upward axis (rad/s).

That is the absolute values, but IMU measures the relative acceleration and angular rate in its coordinate. Why doesn't use the following values?

ax: acceleration in x, i.e. in direction of vehicle front (m/s^2).
ay: acceleration in y, i.e. in direction of vehicle left (m/s^2).
ay: acceleration in z, i.e. in direction of vehicle top (m/s^2).
wx: angular rate around x (rad/s).
wy: angular rate around y (rad/s).
wz: angular rate around z (rad/s).

I'm running this package on Ubuntu 16.04 with python aliased to python3, I ran into this exception message:

Traceback (most recent call last):
  File "/usr/local/bin/kitti2bag", line 12, in <module>
    import tf
  File "/opt/ros/kinetic/lib/python2.7/dist-packages/tf/__init__.py", line 28, in <module>
    from tf2_ros import TransformException as Exception, ConnectivityException, LookupException, ExtrapolationException
  File "/opt/ros/kinetic/lib/python2.7/dist-packages/tf2_ros/__init__.py", line 37, in <module>
    from tf2_py import *
  File "/opt/ros/kinetic/lib/python2.7/dist-packages/tf2_py/__init__.py", line 37, in <module>
    from _tf2 import *
ImportError: No module named '_tf2'

Is there some dependencies I'm missing? It seems to be related to this issue: ros/geometry2#259 but I'm not sure. I've followed the tutorial on tf to install the packages.

Thanks!

@toams789. I obtained one bag file with the 'kitti2bag'. When this bag file was employed to LOAM_velodyne(https://github.com/laboshinl/loam_velodyne), the program failed with an error "failed to find match for field 'intensity' 'failed to find match for field 'ring'". Is there the intensity and ring field in the sensor_msgs::PointCloud2 in the bag file generated by 'kitti2bag'?

Hi,

I generated a bag file of sequence 6 of the KITTI odometry dataset from the raw data recordings using this tool. I then wrote a simple node that uses the rosbag tf to lookup the left camera coordinate pose at each frame and then stores this trajectory. I used the KITTI evaluation tool to plot and evaluate the trajectory.

As you can see, it is pretty close to the ground truth which is expected. However, I would expect it to be exactly the same or very near to it as the paper (Are we ready for autonomous driving? the KITTI vision benchmark suite) says: "The ground truth for visual odometry/SLAM is directly given by the output of the GPS/IMU localization unit projected into the coordinate system of the left camera after rectification."

I did a pose by pose comparison and plotted the translation RMSE (distance between the poses) and for the most part it is close, but there is some significant error at some frames. I haven't been able to determine why there is that error as the ground truth is from the GPS/IMU data directly and the tfs are based off of that.

Do you have any idea what may be the cause?

Thanks,
Vincent

when i run it ,error occured:
kitti2bag -t 2011_09_26 -r 0002 raw_synced .
Traceback (most recent call last):
File "/home/lizj/.local/bin/kitti2bag", line 12, in
import tf
File "/opt/ros/melodic/lib/python2.7/dist-packages/tf/init.py", line 30, in
from tf2_ros import TransformException as Exception, ConnectivityException, LookupException, ExtrapolationException
File "/opt/ros/melodic/lib/python2.7/dist-packages/tf2_ros/init.py", line 38, in
from tf2_py import *
File "/opt/ros/melodic/lib/python2.7/dist-packages/tf2_py/init.py", line 38, in
from ._tf2 import *
ImportError: dynamic module does not define module export function (PyInit__tf2)

Traceback (most recent call last):
File "/usr/local/bin/kitti2bag", line 24, in
from cv_bridge import CvBridge
ImportError: No module named cv_bridge

Hello,
I converted the kitti data to bag file using kitti2bag. When I want to use it in ros and using the following command:
pcl::PointCloudpcl::PointXYZI::Ptr current_sensor_cloud_ptr(new pcl::PointCloudpcl::PointXYZI);
pcl::fromROSMsg(in_sensor_cloud, *current_sensor_cloud_ptr);

it says that "Failed to find match for field 'intensity'.".
When I print names of filed of point cloud data I see that the third filed name is "i". how can I change the name or set the name manually to "intensity"?

on the 113th line of kitti2bag

    velo_datetimes = map(lambda x: datetime.strptime(x[:-4], '%Y-%m-%d %H:%M:%S.%f'), f.readlines())

so if there are empty lines in the timestamps.txt file, this code will fail with:

    ValueError: time data '' does not match format '%Y-%m-%d %H:%M:%S.%f'

And there are indeed empty lines in some kitti datasets, e.g. 2011_09_26_drive_0009_sync

After installing kitti2bag with pip, I noticed that the camera_info width and height are mixed up in the bag file output by the tool. This issue was fixed in the repository, but the latest release (2 years old) doesn't reflect this. Could this be fixed, even just as a one-line 1.5.1 fix?