DLR Holoscene
and DLR SDP
platforms record sensor data into a ZIP file.
Use this repository to convert it into a single rosbag for further analysis and visualization.
Tested on Ubuntu 20.04 and macOS Monterey (note: slow performance on M1 due to architecture mismatch)
- Clone this repository:
git clone https://github.com/borealbikes-dev/combo-rosbag-fuser.git
- Install Docker for your system. Once installed, confirm installation by:
docker -v
On a Mac, you may need to run the Docker application (so that it shows up on the top bar) as well.
In case docker
requires you to run it with sudo
, run the below scripts with sudo
as well.
- Run the main script to download the latest docker image and to create the
input
andoutput
directories.
./run_fuser.sh
-
Move/copy the ZIP file obtained from the Boreal Web Interface into the
input
directory. -
Then run the script again:
./run_fuser.sh
-
See
output
directory for a new.mcap
rosbag. -
Use Foxglove to view its contents.
An unzipped combo_bag looks something like this:
combo_bag_2023-01-25_17-03-15/ # date format: YYYY-MM-DD_HH-MM-SS
├── camera
│ └── usb2 # a directory for each camera source
│ ├── usb-1674662595562-ms-000-minutes.mp4 # filename contains start of video for this camera in milliseconds (since linux epoch)
│ └── usb-1674662595562-ms-001-minutes.mp4 # files are split by 1 minute when recording
└── rosbag
├── metadata.yaml
├── rosbag_0.mcap # rosbag for first minute
└── rosbag_1.mcap # rosbag for second minute