vS-Graphs interfaces with several ROS topics to process sensor data, publish semantic insights, and support modular integrations. The topics are available below:
| Topic | Type | Description |
|---|---|---|
/imu |
sensor_msgs::Imu |
Inertial data for IMU-powered configurations |
/camera/rgb/image_raw |
sensor_msgs::Image |
RGB stream used in RGB-D configurations |
/camera/depth_registered/image_raw |
sensor_msgs::Image |
Aligned depth image stream in RGB-D setups |
/camera/depth/points |
sensor_msgs::PointCloud2 |
Point cloud generated by the depth sensor |
/aruco_marker_publisher/markers |
aruco_msgs::MarkerArray |
Detected fiducial markers from the aruco_ros library |
/camera/color/image_segment |
segmenter_ros::SegmenterDataMsg |
Semantic segmentation results |
/camera/color/image_segment_vis |
segmenter_ros::VSGraphDataMsg |
Visualized semantic segmentation frame |
/voxblox_skeletonizer/sparse_graph |
visualization_msgs::MarkerArray |
Sparse graph of the free-space using vS-Graphs Tools |
/gnn_room_detector |
vs_graphs::VSGraphsAllDetectdetRooms |
GNN-based room segmentation output |
馃挕 To customize input topics and other parameters at launch time, please refer to the 鈿欙笍 vS-Graphs Launch Arguments page for a full list and descriptions.
| Topic | Msg Type | Description |
|---|---|---|
/tf |
tf2_msgs/TFMessage |
Camera transforms in the world frame |
/vs_graphs/camera_pose |
geometry_msgs/PoseStamped |
Camera pose in the world frame |
/vs_graphs/body_odom |
nav_msgs/Odometry |
IMU-body odometry in the world frame |
/vs_graphs/tracking_image |
sensor_msgs/Image |
Image with keypoints and tracking status overlay |
/vs_graphs/tracked_points |
sensor_msgs/PointCloud2 |
Currently tracked SLAM map points |
/vs_graphs/all_points |
sensor_msgs/PointCloud2 |
All SLAM map points |
/vs_graphs/kf_markers |
visualization_msgs/MarkerArray |
Visualization markers for keyframe poses |
/vs_graphs/keyframe_list |
nav_msgs/Path |
Full list of keyframes (poses) |
/vs_graphs/world_frame_pointcloud |
sensor_msgs/PointCloud2 |
Point cloud transformed to the world frame |
/vs_graphs/keyframe_image |
segmenter_ros/VSGraphDataMsg |
Keyframe data with semantic overlays |
/vs_graphs/fiducial_markers |
visualization_msgs/MarkerArray |
Mapped fiducial (e.g., ArUco) markers in the world frame |
/vs_graphs/doorways |
visualization_msgs/MarkerArray |
Mapped doorways in the environment |
/vs_graphs/plane_labels |
visualization_msgs/MarkerArray |
Semantic labels for building components (walls, ground) |
/vs_graphs/building_components |
sensor_msgs/PointCloud2 |
Point clouds of detected building components |
/vs_graphs/segmented_point_clouds |
sensor_msgs/PointCloud2 |
Point cloud after semantic segmentation |
/vs_graphs/freespace_clusters |
sensor_msgs/PointCloud2 |
Clustered free-space regions used for structural analysis |
/vs_graphs/all_mapped_walls |
vs_graphs/VSGraphsAllWallsData |
All wall segments identified (GNN-based room detection) |
/vs_graphs/structural_elements |
visualization_msgs/MarkerArray |
Detected structural elements (rooms and corridors) |
/vs_graphs/camera_pose_vis |
visualization_msgs/MarkerArray |
Visualization markers for current camera pose |