This repository contains starting points and configuration for some project ideas, some involving processing real data and some involving Gazebo simulations. The bag files to use for these sample projects can be found here: https://1drv.ms/f/s!Ar6id-4c-fy3yDMMXjcEX7lZRzkx?e=J0cv1t.
The dataset_playback
package contains a launch file with supporting configuration to play back the data from the following bag files:
road_data_sample.bag
road_data_sample_with_labels.bag
pedestrian.bag
pedestrian_with_labels.bag
stereo_camera_example.bag
dataset_playback.launch
(dataset_playback/launch/dataset_playback.launch) runs the necessary nodes to process the raw data from the bag and set up the TF frame tree to align all the sensor data with the vehicle body frame. It also uses the RTK GNSS position to update a TF frame from a central map frame to the vehicle body frame.
Use real data from a 3D LIDAR sensor to construct a map of an environment, then use the map for localizing the vehicle as an alternative to GPS.
For more details, see the project-specific README
Process data from multiple sensors to perceive better information about the environment than is possible with a single sensor.
For more details, see the project-specific README
Detect a traffic light and determine its state in Gazebo, then use that information to control the car to continue through or stop at the intersection.
For more details, see the project-specific README
Drive around an oval track with other vehicles. There are some two-lane sections that could be used for passing...
For more details, see the project-specific README