RoboDine


Vision and robotics project for monocular robot-arm pose estimation, foreign-object handling, and Agile collaboration

Documentation ROS2 Vision DBSCAN

RoboDine is a vision and robotics project that trained a YOLO pose model with data preprocessing and augmentation to reduce pre-grasp pose-estimation errors for a robot arm with a monocular 2D camera. It improved 6-DoF pose-estimation accuracy by 14%, implemented Hough Circles and DBSCAN-based logic for foreign-object scenarios in the loading area, and used GitHub, Jira, and Confluence in an Agile workflow.

Highlights

  • Trained a YOLO pose model with data preprocessing and augmentation to reduce pre-grasp pose-estimation errors for a robot arm with a monocular 2D camera, improving 6-DoF pose-estimation accuracy by 14%.
  • Implemented Hough Circles and DBSCAN-based logic to prepare for scenarios where foreign objects cause issues in the loading area.
  • Collaborated through GitHub, Jira, and Confluence while following an Agile development process.

Demo Videos

RoboDine integrated operation demo
Vision and avoidance pipeline demo

Repository

git clone https://github.com/addinedu-roscamp-4th/roscamp-repo-2.git
cd roscamp-repo-2