AR Tag Detection

ROS package and mission assets for AR marker bundle detection in the CORODRO project. This repository contains the launch files, bundle definitions, and printable marker utilities used to localize mission objects and shared reference frames from RGB camera observations.

Why This Repo Exists

The rover and drone needed a simple, reproducible way to recognize mission targets and align local coordinate frames. This repository packages the AR-tag assets that were used for that workflow so students can reproduce the setup without digging through mission code first.

Added Value

• Ready-to-use marker bundle XML files for the IGLUNA boxes and points of interest.
• Launch files for running ar_track_alvar with the Intel RealSense D435 topics used in the project.
• Utilities for regenerating bundle definitions and printable marker sheets.

Algorithms Used

• ARTrack Alvar multi-marker bundle detection.
• Static transform publication for camera-to-robot frame alignment.
• Bundle-based tag grouping so a box can still be recognized when only part of it is visible.

Repository Map

• launch/: canonical entry points for camera-based AR tracking.
• bundles/: mission bundle definitions for points of interest and boxes.
• bundles/final_bundles/: per-box bundle XML files plus the helper used to generate them.
• src/Marker_igluna/: marker generation and printable assets.
• src/: legacy workspace mirror kept for compatibility with the original student workspace layout.

Quick Start

1. Add this package to a ROS 1 catkin workspace that already has ar_track_alvar installed.
2. Connect the D435 topics or remap them to your own RGB camera topics.
3. Start with launch/ar_track_D435_cam.launch for single-marker detection or launch/ar_track_D435_cam_multitag.launch for bundle detection.
4. Use launch/ar_track_D435_cam_multitag_box_world.launch when you want bundle detection in the world frame with the provided static transform.
5. Adjust marker size, output frame, bundle file, and static-transform arguments according to your setup.

Recommended Entry Points

• launch/ar_track_D435_cam.launch: simplest single-marker detector.
• launch/ar_track_D435_cam_multitag.launch: best starting point for bundle detection.
• launch/ar_track_D435_cam_multitag_box_world.launch: box-focused bundle launch with a world-frame output.
• launch/multitag_23_box.launch: launches the full final-box bundle list used for the complete mission marker set.

Notes For Students

• The repository name is legacy. If you reuse this outside the original workspace, ar_tag_detection is the clearer convention.
• Prefer the top-level launch/ and bundles/ folders. The mirrored files under src/ are legacy copies from the original workspace structure.
• Some launch files still reflect the original mission frame names and should be adapted to your robot.
• The core detector comes from ar_track_alvar; this repo is mainly the mission-specific configuration layer.

License

This repository includes a BSD 2-Clause license file from the original package authoring work.