Using YOLO and roboflow to track poker cards
The main objective of this sub-project is to develop a tracking model that accurately tracks poker cards (flop, turn and river). Another personal objective of this project is to learn about the new YOLOv8 API and more generally how tracking algorithms work.
The Playing Cards dataset is a collection of synthetically generated cards blended into various types of backgrounds. You will be able to perform object detection to detect both number and suit of the cards. The dataset contains 10100 images that were preprocessed and augmented resulting in a total of 24240 which were consequently split into 21210-2020-1010 for training, validation and testing respectively.
- Auto-Orient: Applied
- Resize: Stretch to 640x640
Outputs per training example: 3
- Flip: Horizontal, Vertical
- 90° Rotate: Clockwise, Counter-Clockwise, Upside Down
- Crop: 0% Minimum Zoom, 15% Maximum Zoom
- Rotation: Between -10° and +10°
- Shear: ±2° Horizontal, ±2° Vertical
- Grayscale: Apply to 10% of images
- Hue: Between -25° and +25°
- Saturation: Between -25% and +25%
- Brightness: Between -25% and +25%
- Exposure: Between -15% and +15%
- Blur: Up to 1.75px
- Noise: Up to 2% of pixels
- Cutout: 5 boxes with 2% size each
Dataset: [roboflow playing card dataset](https://github.com/ultralytics/yolov5https://universe.roboflow.com/augmented-startups/playing-cards-ow27d/dataset/4]
- An IP camera will be installed on top of the poker table or at a favourable angle for detecting cards on the borad. The model should be running and continiously detecting cards. As the cards are opened (flor-turn-river), the predicted labels should be added to the database.
- With the RFID readers installed to read the players cards, a complete and live view of the action should be running.