y', docker image 'udacity-reader'...
Unable to find image 'udacity-reader:latest' locally
docker: Error response from daemon: pull access denied for udacity-reader, repository does not exist or may require 'docker login': denied: requested access to the resource is denied.
See 'docker run --help'.

Hello, I'm getting the next error while trying to run sudo ./build.sh:

I don't know why this is happening since I'm using python 3.6.9 and already removed the 2.7 version.

Please help!

Hey,

Thanks a lot for the data set.

I am having a issue, when I try to use the bagdump.sh it is giving an error
File "/usr/local/lib/python2.7/dist-packages/pandas/compat/init.py", line 421, in
raise ImportError('dateutil 2.5.0 is the minimum required version')
ImportError: dateutil 2.5.0 is the minimum required version

Do you have any idea why this might be happening.

Thanks

I am getting manifest error when trying to install

Error response from daemon: manifest for timeshi1995/udacity-driving-reader:latest not found: manifest unknown: manifest unknown

Hi,

Thanks for the dataset reader code.

I'm interested in extracting point correspondence between consecutive images in the dataset. I'm interested in 3D points present in sensor_msgs/PointCloud2 and the extrinsic camera parameters in either sensor_msgs/JointState or tf2_msgs/TFMessage.

Could you please provide a more generic code that can extract this data into csv files too?

Many thanks.

Hi!
I tried to run ./run-bagdump.sh -i /data -o /output , but got the error
./run.sh: line 32: docker: command not found
I set my folders follow this image,
please help me!!

File "/usr/local/lib/python2.7/dist-packages/pandas/compat/init.py", line 421, in
raise ImportError('dateutil 2.5.0 is the minimum required version')
ImportError: dateutil 2.5.0 is the minimum required version

I have dateutil version 2.7 yet its giving the error.

Hi thanks for your nice script.

Howeer, I'm trying to run the script in windows 10, any structions on how to do so? How about directly run the ./build.sh in powershell? And then run the ./run-bagdump.sh command?

In addition, I'm new to this docker thing. Can you please briefly specify how to "Build the docker container manually"

Or a simpler way, may I ask is there any way I can download the extracted image files and csv files?

Hello there,
was wondering if there are many machine specs required to run the script? because I'm trying to run it on a large dataset (80 GBs) and it's stuck at processing dataset, never making any progress or writing any images and it causes my ubuntu to freeze.
Thanks

There were several issues I faced while trying to extract the data from .bag files:
Docker file:
When trying to run as is:

• pandas throws errors stating numpy version is 11.5.0, while >=12.0.0 is needed.
• Similarly, several other PyPi packages that are already installed in ros:kinetic-vision image are left as older versions since there's no --upgrade flag at the end of pip commands(ex. PyYaml, numpy, matplotlib, ipykernel, python-dateutil), which causes problems.

Run.sh:

• Doesn't do anything for me since Docker expects absolute paths for OUTPUT_DIR and INPUT_DIR, while given paths are relative to current directory.

After fixing the above problems, everything works as expected. I think Docker file and Run.sh need small changes. I have opened a pull request, please have a look.

The script doesn't extract Velodyne packets/data from the .bag file. While the yaml shows inclusion of the following topic: /velodyne_packets, the topic is not accessed via the script.
Can you please add that?

Hello,

What are the units of the sensor data?
Specifically, I want to use steering angle and speed. Are those in radians and kmh?
Moreover, what are the time units? It seems like it uses [sec*1e-9], but I want to make sure.

Btw, thanks for the great tool, it saved me a lot of time.

#12
The absolute path also don't work!
Can you help me?
Thank you very much!

Built the docker environment manually and tried to run python bagdump.py.

Got the error No handler could be found for logger "rosout"

Had to run source /opt/ros/<distro>/setup.bash in container's bash before running python bagdump.py.

Hope this helps ppl who are facing the same issue

iam using window 10 home which is not compatable for installing docker ,suggest any alternative

I am new to docker and I have tried reading and watching tutorials on the internet for a while, but it's not helpful. I wonder how I should run the script files within the Container. Am I supposed to copy this reader directory to a container so that I can run the script files ?

this is that in my terminal, could someone help me please...

root@MONTO:/mnt/j/ucity/udacity-driving-reader-master# ./run-bagdump.sh -i /mnt/j/ucity -o /mnt/j/ucity/data
Running 'python script/bagdump.py' with input dir '/mnt/j/ucity', output dir '/mnt/j/ucity/data', docker image 'udacity-reader'...
python: can't open file 'script/bagdump.py': [Errno 2] No such file or directory

thanks!

Even after modifying the docker file with the comment in issue #24 and pull request #28, the terminal gives me the following error:

29 | >>> RUN pip install
30 | >>> scipy==1.2.3
31 | >>> pandas==0.24.2
32 | >>> jupyter==1.0.0
33 |

ERROR: failed to solve: process "/bin/sh -c pip install scipy==1.2.3 pandas==0.24.2 jupyter==1.0.0" did not complete successfully: exit code: 1

Thank you very much for this script. Just a heads up that for small data sets, sub 6000 images, the user will encounter an error per below. Its easy to debug, but a warning might help future users. Thanks again.

File "script/bag2tf.py", line 475, in main
splits=split_list, center_only=center_only, debug_print=debug_print)
File "script/bag2tf.py", line 207, in init
writer = ShardWriter(self._outdir, s[0], scaled_images, max_num_shards=scaled_shards)
File "script/bag2tf.py", line 135, in init
self.num_entries_per_shard = num_entries // max_num_shards
ZeroDivisionError: float divmod()

Hi!

I used this reader to convert the dataset for Udacity challenge 2 from .bag files to images and .csv files, but the timestamps in the break.csv do not correspond either to the ones in the interpolated.csv nor to the images filenames (link to the 2 files). This also happens for timestamps in the gear.csv or gps.csv. As it was recommended in the repository readme, I used this script.

Do you know why this may happen? or how can I correlate the timestamps in brake.csv, for instance with the images?

Thanks,
Miruna

docker: Error response from daemon: pull access denied for udacity-reader, repository does not exist or may require 'docker login'.