rosbag2video

rosbag2video.py
rosbag to video file conversion tool
by Maximilian Laiacker 2025
[email protected]

with contributions from
David Silva 2025,
Abel Gabor 2019,
Bey Hao Yun 2024,
[email protected],
[email protected]
[email protected]

Install

Build docker images using the commands below:

1. Download rosbag2video:

install python3

cd $HOME

git clone https://github.com/mlaiacker/rosbag2video

install dependencies

sudo apt install ffmpeg
pip install rosbags
pip install opencv-python

Mostly tested with bags containing

msg_type: sensor_msgs/msg/CompressedImage msg_encoding: jpeg

Or use docker if you want to:

2. Build docker image for running ROS 1 rosbag2video.py:

docker build -f Dockerfile.ros1 -t rosbag2video:noetic .

3. Build docker image for running ROS 2 ros2bag2video.py:

docker build -f Dockerfile.ros2 -t rosbag2video:humble .

Usage

By default it will extract all compressed image topics inside the bag directory with the name of .mp4 ('/' inside the topic name will be replaced by '_') as mjpeg encoded video with 30fps.

usage: rosbag2video [-h] [-v] [-r RATE] [-t TOPIC] [-o OFILE] [--save_images] [--frames FRAMES] rosbag [rosbag ...]

Convert ROS bag (1/2) to video using ffmpeg.

positional arguments:
  rosbag                Input File(s)

options:
  -h, --help            show this help message and exit
  -v, --verbose         Run rosbag2video script in verbose mode.
  -r RATE, --rate RATE  Video framerate
  -t TOPIC, --topic TOPIC
                        Topic Name
  -o OFILE, --ofile OFILE
                        Output File
  --save_images         Boolean flag for saving extracted .png frames in frames/
  --frames FRAMES       Limit the number of frames to export

ROS 1

docker run -it --rm \
    --name rosbag2video_c \
    -v .:/rosbag2video_workspace \
  rosbag2video:noetic bash

source /opt/ros/noetic/setup.bash

python3 rosbag2video.py -t <topic_name> -i <bag_file_name> -o <output_video_file_name>
# Eg. python3 rosbag2video.py -t /cam0/image_raw -i ar_tracking_1.bag -o myvideo.mp4

ROS 2

docker run -it --rm \
    --name rosbag2video_c \
    -v .:/rosbag2video_workspace \
  rosbag2video:humble bash

source /opt/ros/humble/setup.bash

Example: extract all image topics form one ore more rosbag2 directories or bags

python3 rosbag2video.py <bag_folder_name>
# Eg. python3 rosbag2video.py rosbag2_2024_10_11-19_45_28 -o myvideo.mp4

Example: extract a specific topic

python3 rosbag2video.py -t <topic_name> -o <output_video_file_name> <bag_folder_name>
# Eg. python3 rosbag2video.py -t /cam0/image_raw -o myvideo.mp4 rosbag2_2024_10_11-19_45_28