stihl_visualization

ROS2 package providing lightweight visualization utilities for the Stihl mobile platform.

This package contains ROS2 nodes and utilities for visualizing a robot mesh on a common map frame, publishing a polygon that demarks an operational area, rendering detected weeds as markers, and publishing prerecorded points as a nav_msgs/Path message for inspection in Foxglove Studio or RViz.

Contents

• Nodes: - visualization — reads a polygon (lat/lon) from config/stihl.yaml, projects coordinates to a local tangent plane in meters, publishes a polygon Marker, and places a robot mesh using live GNSS fixes (sensor_msgs/NavSatFix). - heatmap_plotter — maintains a simple grid/heatmap of weed detections and publishes an OccupancyGrid and text MarkerArray for non-empty cells.
• Config: - config/stihl.yaml — polygon coordinates (latitude, longitude) used by polygon_visualization. - other example configs in config/ used by supporting nodes (e.g. heatmap settings).

Installation

Prerequisites:

• ROS2 (tested with Humble; follow official installation docs for your distro)
• Python dependencies used by package nodes (install via rosdep below)

Typical setup steps (from workspace root):

# source ROS2
source /opt/ros/humble/setup.bash

cd ~/stihl_ws
rosdep install -i --from-path src --rosdistro humble -y
colcon build
source install/setup.bash

If you add or change Python dependencies, ensure they are installed in the environment used for ROS2 (system Python, venv, or colcon's install).

Usage

Start the visualizations using the provided launch file (example):

source install/setup.bash
ros2 launch stihl_visualization visualization.launch.py

You can also run the console nodes directly (installed as console scripts by setup.py):

ros2 run stihl_visualization visualization
ros2 run stihl_visualization heatmap_plotter

movement_test.py is a convenience/demo script included in the package source — it may be executed directly for quick local debugging (it is not registered as a console script by default).

Nodes, Topics and Messages

• visualization node - Subscribes: /ublox_gps_node/fix — sensor_msgs/NavSatFix (optional; used to place the robot mesh at live GNSS-derived local positions) - Publishes: - /polygon_marker — visualization_msgs/Marker (LINE_STRIP) in frame map - /robot_marker — visualization_msgs/Marker (MESH_RESOURCE) for the robot mesh - /ground_marker — visualization_msgs/Marker (CUBE) for a ground plane

• heatmap_plotter node - Subscribes: GPS and detection topics (configured via node config) - Publishes: an nav_msgs/OccupancyGrid heatmap and visualization_msgs/MarkerArray for text markers

All visualization outputs use the map frame for consistent overlay in Foxglove Studio or RViz.

Coordinate conversions

• Geodetic → local planar meters (used by visualization):
  • Latitude/longitude pairs in config/stihl.yaml are projected to a local tangent plane using an Azimuthal Equidistant (AEQD) projection centered on the first coordinate in the YAML file. The result is in meters (local ENU-like coordinates) so the polygon lines and robot mesh align in the map frame.

Configuration

• config/stihl.yaml — must contain polygon_coordinates as a list of [lat, lon] pairs. The first pair is used as the projection origin.

Other config files in config/ provide example parameters for heatmap sizing and behavior; review them to tune grid extent, resolution and topics.

Developer notes / API

• stihl_visualization/visualization.py — VisualizationNode implementation. Reads config/stihl.yaml, computes a local projection using pyproj, publishes polygon Marker and robot mesh placements from live GNSS fixes.
• stihl_visualization/heatmap_plotter.py — HeatmapPlotter implementation. Maintains a grid/heatmap of weed detections and publishes nav_msgs/OccupancyGrid plus marker overlays.
• meshes/ — place mesh assets here (e.g., stihl.stl). URIs use the package scheme, e.g. package://stihl_visualization/meshes/stihl.stl.

Troubleshooting and tips

• If markers or meshes are not visible in Foxglove Studio or RViz, ensure the workspace is properly built and sourced so package:// URIs resolve:

source install/setup.bash

• Missing YAML: polygon_visualization will raise an error if config/stihl.yaml is absent or lacks polygon_coordinates — ensure the file exists and contains at least two coordinates.
• Mesh scale: markers set mesh scale factors small by default; tune marker.scale values to fit the scene visually.