Laser Runner Cutter

Laser Runner Cutter is a project by the California Strawberry Commission for automated detection and cutting of strawberry runners (also known as stolons).

Packages

ROS2/

• runner_cutter_control: node for laser runner cutter control and automation.
• runner_cutter_control_interfaces: custom interface definitions used in runner_cutter_control.
• laser_control: node and client for controlling laser DACs. Currently supports Helios and Ether Dream DACs.
• laser_control_interfaces: custom interface definitions used in laser_control.
• camera_control: node and client for capturing camera frames and running runner and laser detection. Currently supports RealSense cameras.
• camera_control_interfaces: custom interface definitions used in camera_control.
• common_interfaces: shared interface definitions
• rqt_laser_runner_cutter: rqt plugin for visualization and monitoring of relevant ROS2 topics.

ML/

• runner_segmentation_model: ML pipeline and models for instance segmentation of runners
• laser_detection_model: ML pipeline and models for object detection of lasers
• runner_mask_instancing_model: ML pipeline and models for instace segmentation of semantic segmented runner masks. Goal: given a single binary mask that represents runners, segment it into separate instances of runners.

Apps/

• runner-cutter-app: Web app for laser runner cutter control and automation.
• runner_image_capture_amiga_kivy: An application deployable to the Amiga brain, written using Kivy, for capturing runner images in the field.
• runner_image_capture_amiga_react: An application deployable to the Amiga brain, written using React and FastAPI, for capturing runner images in the field.

Tools/

• simulator: Unity-based simulator.

Setup and run

1. Setup and run ROS2 nodes: see ros2/README.md

2. Setup and run Runner Cutter App: see apps/runner-cutter-app/README.md