MohammedHemed/Chessboard-digital-images_with_fen

Chessboard Detection Dataset

This dataset consists of a total of 64,386 chessboard images and corresponding YOLO-format label files.

Dataset Breakdown

• Images: 64,386 total

  • train: 57,928
  • val: 6,458

• Labels: 64,386 total (one .txt per image)

  • train: 57,928
  • val: 6,458

Each label file contains bounding boxes for the pieces on the board using YOLO format. The dataset includes 12 classes:

• 6 white pieces
• 6 black pieces

---

Data Collection & Annotation

The dataset was generated using chess game data from the Lichess platform, which provides a massive monthly collection of games in PGN format. Each game includes a FEN string for every move, describing the position of all pieces on the board.

We used:

• The python-chess API to convert FEN strings into rendered chessboard images.
• A custom script to divide the board into 8×8 squares and extract object annotations from each FEN.
• These annotations were then converted into YOLO-format .txt files for training object detection models.

---

Use Cases

This dataset is ideal for:

• Training object detection models (YOLOv5, YOLOv8, etc.)
• Detecting individual chess pieces on a board
• Converting board images back into digital game state (FEN)

---

License

This dataset is distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.