This repository contains the hardware design, software implementation and thesis for an automated system designed to recognize and classify trading cards (specifically Magic: The Gathering).
The project was developed to design and build a simple robot capable of feeding trading cards one at a time under a camera. The captured images are then processed using computer vision and machine learning techniques to identify and classify each card.
The system is divided into two main functional parts:
- Hardware: An automated card-feeding mechanism driven by stepper motor and controlled via an Arduino board.
- Software: A Python-based recognition pipeline using machine learning (ArcFace) to identify the scanned cards based on a constructed dataset.
/hw— hardware specifications. Contains photos, wiring diagrams and the final selection of components for the robot./sw— source code for the project. Includes Arduino firmware for the robotic feeder and Python scripts for the machine learning and image processing pipeline./scripts— various utility scripts, including tools for downloading and preparing image datasets./sources— collected research materials, links to scientific articles (papers) and tutorials on relevant algorithms./thesis— LaTeX source files of the thesis document./presentation— slides and materials used for the thesis presentation/defense.
This project is licensed under the GNU Affero General Public License v3.0.