KITTI to ROS2 Bag

The KITTI dataset is one of the most widely used benchmarks in autonomous driving research, but working with it in a ROS2 development environment requires tedious manual conversion. This package automates the full conversion — reading synced camera images, LiDAR point clouds, IMU, and GPS data from the KITTI directory structure and writing them into a ROS2 bag file with correct timestamps and topic names.

What it does

• Reads the KITTI Synced & Rectified dataset (camera, Velodyne, oxts)
• Publishes sensor_msgs/Image, sensor_msgs/PointCloud2, sensor_msgs/Imu, and sensor_msgs/NavSatFix messages
• Preserves original KITTI timestamps across all topics
• Outputs a valid ROS2 bag playable with ros2 bag play

Stack: Python 3, rclpy, rosbag2_py, sensor_msgs