rosbag-util 1.0.4

Utilities for extracting synchronized frames from ROS bag files.

rosbag-util

Extract synchronized point cloud frames (and optional GPS) from one or more ROS bags.

Install

pip install -e .

Prerequisites

• ROS Python environment with rosbag and sensor_msgs available (e.g. ROS Noetic), and source your ROS setup.

Usage

Single car:

rosbag-extract \
  --bags "/path/to/car1.bag" \
  --pc-topic "/perception/lidar/concated_points_cloud" \
  --gps-topic "/location/best_position" \
  --out "cooperative/car1" \
  --binary

Batch undistort images:

python -m rosbag_util.undistort_images \
  --input ./imgs \
  --output ./undist \
  --cam 3 \
  --model pinhole \
  --camera-config ./configs/camera_config.json

Using a config file:

python -m rosbag_util.undistort_images --config ./configs/undistort_config.json

Multi car (car1 as main):

rosbag-extract \
  --bags "/path/to/car1.bag" "/path/to/car2.bag" \
  --main 0 \
  --pc-topic "/perception/lidar/concated_points_cloud" \
  --gps-topic "/location/best_position" \
  --out "cooperative/scene1" \
  --max-dt 0.300 \
  --save-workers 8 \
  --index-threads 2 \
  --binary

Config file (JSON):

{
  "bags": ["/path/to/car1.bag", "/path/to/car2.bag"],
  "main": 0,
  "pc_topic": "/perception/lidar/concated_points_cloud",
  "gps_topic": "/location/best_position",
  "out": "cooperative/scene1",
  "max_dt": 0.3,
  "binary": true,
  "save_workers": 8,
  "index_threads": 2
}

rosbag-extract --config config.json