surawut/chessboard-dataset-yolo

Chessboard Corner Pose Dataset ♟️📐

1. Dataset Description

This dataset is designed for Keypoint Detection (Pose Estimation) tasks, specifically to detect the 4 semantic corners of a chessboard from various camera angles.

It is intended to be used as Phase 1 of a Chess Move Tracking pipeline. The detected keypoints are used to calculate the Homography matrix for Perspective Warping (straightening the board).

• Task: Keypoint Detection (YOLO-Pose)
• Format: YOLOv8 / YOLO11 Pose Format
• Input: Raw video frames (rotated, tilted, occluded)
• Output: 4 Keypoints (a1, h1, a8, h8)

2. Dataset Structure

Keypoints Definition

The model is trained to detect 1 Class (Chessboard) with 4 Keypoints:

1. Index 0: a1 (Bottom-Left in White perspective)
2. Index 1: h1 (Bottom-Right in White perspective)
3. Index 2: a8 (Top-Left in White perspective)
4. Index 3: h8 (Top-Right in White perspective)

Data Splits

Split	Images	Description
Train	3390	Extracted from Various Chess Play Videos with augmentations
Valid	290	Extracted from Various Chess Play Videos (Unseen angles)