No version for distro humble showing github. Known supported distros are highlighted in the buttons above.

carla_perception package from wato_monorepo repo

action_bringup behaviour_planning global_planning local_planning model_predictive_control wato_trajectory_msgs infrastructure_bringup tf_static_relay vision_msgs_markers wato_lifecycle_manager ackermann_mux camera_calib lidar_lidar_calib can_state_estimator dummy_ackermann eve_description interfacing_bringup joystick_interfacing oscc_interfacing pid_control roscco_msg sensor_interfacing camera_object_detection depth_estimation feature_extraction patchworkpp patchworkpp_vendor perception_bringup perception_utils bytetrack_cpp_vendor track_viz_2d tracking traffic_light_detection carla_bringup carla_common carla_control carla_lifecycle carla_localization carla_msgs carla_perception carla_pygame carla_ros_bridge carla_sample_description carla_scenarios carla_teleop simulation_bringup wato_test behaviortree_ros2_vendor behaviour behaviour_markers behaviour_msgs costmap costmap_markers costmap_msgs lanelet_markers lanelet_msgs lidar_localization localization prediction simple_prediction world_model world_model_markers world_model_msgs world_modeling_bringup

ROS Distro
github

API Docs
Browse Code

Package Summary

Version	1.0.0
License	Apache-2.0
Build type	AMENT_PYTHON
Use	RECOMMENDED

Repository Summary

Description	Dockerized ROS2 stack for the WATonomous Autonomous Driving Software Pipeline
Checkout URI	https://github.com/watonomous/wato_monorepo.git
VCS Type	git
VCS Version	main
Last Updated	2026-02-24
Dev Status	UNKNOWN
Released	UNRELEASED
Contributing	Help Wanted (-) Good First Issues (-) Pull Requests to Review (-)

Package Description

Sensor bridge nodes (camera, lidar, bbox)

Maintainers

WATonomous

Authors

No additional authors.

src/simulation/carla_ros_bridge/carla_perception/README.md

carla_perception

Sensor data publishers for CARLA simulation.

These nodes spawn virtual sensors attached to the ego vehicle in CARLA and publish their data as ROS messages. Sensor positions are read from TF (published by robot_state_publisher from URDF) at activation time, so the URDF frame names must match the configured frame_id parameters.

Nodes

camera_publisher

Spawns RGB cameras in CARLA and publishes images. Supports multiple cameras configured via the camera_names parameter. Each camera’s resolution, FOV, and intrinsics are configured via namespaced parameters (e.g., front_camera.image_width).

Camera intrinsics (K, P matrices) are published alongside images for use with image processing pipelines.

ros2 run carla_perception camera_publisher

Publications: <camera_name>/image_raw (sensor_msgs/Image), <camera_name>/camera_info (sensor_msgs/CameraInfo)

Parameters:

Parameter	Type	Default	Description
`carla_host`	string	`localhost`	CARLA server hostname
`carla_port`	int	`2000`	CARLA server port
`carla_timeout`	double	`10.0`	Connection timeout in seconds
`role_name`	string	`ego_vehicle`	Role name of the ego vehicle to attach sensors to
`camera_names`	string[]	`['camera']`	List of camera sensor names to spawn

Per-camera parameters (replace <name> with camera name):

Parameter	Type	Default	Description
`<name>.frame_id`	string	required	TF frame for camera position
`<name>.image_width`	int	required	Image width in pixels
`<name>.image_height`	int	required	Image height in pixels
`<name>.fov`	double	`90.0`	Horizontal field of view in degrees
`<name>.optical_frame`	bool	`true`	Whether frame_id uses optical convention (Z-forward)
`<name>.camera_matrix.data`	double[]		3x3 camera intrinsic matrix (row-major)
`<name>.distortion_coefficients.data`	double[]		Distortion coefficients
`<name>.projection_matrix.data`	double[]		3x4 projection matrix (row-major)

lidar_publisher

Spawns LiDAR sensors in CARLA and publishes point clouds. Supports multiple LiDARs configured via the lidar_names parameter. Each LiDAR’s range, channels, rotation frequency, and FOV are configured via namespaced parameters.

Point clouds are accumulated over a full rotation before publishing, ensuring each message contains a complete 360° scan (or the configured horizontal FOV).

ros2 run carla_perception lidar_publisher

Publications: <lidar_name>/points (sensor_msgs/PointCloud2)

Parameters:

Parameter	Type	Default	Description
`carla_host`	string	`localhost`	CARLA server hostname
`carla_port`	int	`2000`	CARLA server port
`carla_timeout`	double	`10.0`	Connection timeout in seconds
`role_name`	string	`ego_vehicle`	Role name of the ego vehicle to attach sensors to
`lidar_names`	string[]	`['lidar']`	List of LiDAR sensor names to spawn

Per-LiDAR parameters (replace <name> with LiDAR name):

Parameter	Type	Default	Description
`<name>.frame_id`	string	required	TF frame for LiDAR position
`<name>.range`	double	required	Maximum detection range in meters
`<name>.rotation_frequency`	double	required	Scan rate in Hz
`<name>.points_per_channel`	int	required	Points per channel per rotation
`<name>.channels`	int	`32`	Number of laser channels
`<name>.horizontal_fov`	double	`360.0`	Horizontal field of view in degrees
`<name>.upper_fov`	double	`10.0`	Upper vertical FOV limit in degrees
`<name>.lower_fov`	double	`-30.0`	Lower vertical FOV limit in degrees
`<name>.noise_stddev`	double	`0.0`	Distance noise standard deviation
`<name>.dropoff_general_rate`	double	`0.0`	Random point dropout rate (0.0-1.0)

bbox_publisher

Publishes ground truth 3D bounding boxes for all vehicles and pedestrians in the CARLA world. Useful for perception algorithm development and validation.

ros2 run carla_perception bbox_publisher

Publications: detections_3d (vision_msgs/Detection3DArray), tracked_detections_3d (vision_msgs/Detection3DArray)

ObjectHypothesisWithPose format — each Detection3D.results list is populated as follows:

Object type	`results[0]`	`results[1+]`
Vehicle	`class_id="vehicle"`, `score=1.0`	One entry per active signal: `"left_blinker"`, `"right_blinker"`, `"brake"`, `"reverse"`
Pedestrian	`class_id="pedestrian"`, `score=1.0`	—
Traffic light	`class_id="traffic_light"`, `score=1.0`	`class_id="red"/"yellow"/"green"/"unknown"`, `score=1.0`

Parameters:

File truncated at 100 lines see the full file

CHANGELOG

No CHANGELOG found.

Package Dependencies

Deps	Name
	ament_python
	python3-carla
	rclpy
	sensor_msgs
	vision_msgs
	geometry_msgs
	lifecycle_msgs
	tf2_ros
	tf2_geometry_msgs
	carla_common

System Dependencies

Name
python3-numpy
python3-pytest

Dependant Packages

Name	Deps
carla_bringup
carla_ros_bridge

Launch files

No launch files found

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged `carla_perception` at Robotics Stack Exchange

API Docs Browse Code

No version for distro jazzy showing github. Known supported distros are highlighted in the buttons above.