A robotic system that creates pixel art using a Franka Panda robot arm with colored stamps. The robot autonomously picks up colored stamps, dabs them on ink pads, and stamps pixels on a canvas to recreate digital images in the physical world.
This project implements an end-to-end pipeline for robotic pixel art creation:
- Computer Vision: Processes input images and converts them to robot-executable pixel coordinates
- Task Planning: Optimizes stamp usage and movement paths using Traveling Salesman Problem (TSP) algorithms
- Robot Control: Precise manipulation of a Franka Panda arm for stamp handling and pixel placement
- ROS Integration: Real-time pose estimation and coordination between vision and control systems
Input Image → Image Processing → Pixel Extraction → Task Planning → Robot Execution
- Vision Pipeline: ArUco marker detection for stamp pad localization
- Path Optimization: Christofides algorithm for efficient movement
- Safety Systems: Workspace bounds and collision avoidance
- Robot: Franka Panda 7-DOF arm
- Camera: RGB camera for ArUco detection
- Stamps: 4 colored stamps (Red, Green, Blue, Black)
- Ink Pads: Corresponding colored ink pads
- Canvas: White paper or canvas surface
- Markers: ArUco markers for pose estimation