This kit contains utilities to convert rosbags to KITTI format, involving LiDARs, images, GPS and IMU information.

This is adapted from the VBR dataset devkit. All credits to the original authors of the VBR dataset.

After cloning the repository, install the python package by running:

cd python
pip install .

The sequences are provided in ROS1 format. We offer a convenient tool to change representation if you prefer working on a different format like in KITTI dataset. You can see the supported formats by typing:

rosbag_converter --help

To convert a bag or a sequence of bags, type:

rosbag_converter <desired_format> <input_directory/input_bag> <output_directory>

for instance, we could convert the campus_train0_00 sequence to kitti format as follows:

rosbag_converter kitti ./rosbag/campus_train0_00.bag ./extracted_data/

Note that the input can also be a folder containing multiple bags. The converter will merge them and ouput into a single folder.

Besides, if the RGB images are already debayered, please set the --no-rgb-conversion flag, for example:

rosbag_converter --no-rgb-conversion kitti ./rosbag/campus_train0_00.bag ./extracted_data/

You can expect the following result:

data
  - your_output_dir
    - camera_left
      - timestamps.txt
      - data
        - 0000000000.png
        - 0000000001.png
        - ...
    - camera_right
      - timestamps.txt
      - data
        - 0000000000.png
        - 0000000001.png
        - ...
    - ouster_points
      - timestamps.txt
      - data
        - 0000000000.ply
        - 0000000001.ply
        - ...
    - ... 

The extracted data can be then used as the input to systems such as PINGS together with the provided calibration information.