• Overview
• Features
• Usage
  • Training the Model
    • Download the Playing Cards Image Dataset
    • Organize Dataset in Google Drive
    • Train the Model
    • Retrieve and Download the Best Model
  • Making Real-time Predictions
    • Prerequisites
    • Create a Virtual Environment
    • Clone the Repository
    • Move and Rename the Trained Model
    • Install Necessary Packages
    • Download CUDA Toolkit
    • Run the Script
• References

This repository contains code for training a YOLOv8 model on a large dataset of playing cards to create a poker hand detection system using Python. It includes a setup for using a webcam live feed to make real-time predictions and identify poker hands.

• Real-time detection of poker hands using a webcam.
• Pre-trained YOLOv8 model for quick setup.
• Easy-to-follow training process for custom datasets.
• Integration with CUDA for accelerated performance.

You can skip the training steps and use the pre-trained model provided in the model/ directory.

Download the dataset from the following link:

https://universe.roboflow.com/augmented-startups/playing-cards-ow27d/dataset/4/download/yolov8

Add the dataset to your Google Drive with the following hierarchy:

Download and run the training notebook:

https://github.com/mrkrisgee/poker_hand_detection/tree/main/train

Once training is complete, download the "best" model from:

/runs/detect/train/weights/best.pt

Ensure you have Anaconda installed on your system. Anaconda simplifies package management and deployment.

Create and activate a new conda environment by running the following commands in your terminal:

conda create -n yolov8
conda activate yolov8

Clone this repository to your local machine and navigate into the project directory:

git clone https://github.com/mrkrisgee/poker_hand_detection.git
cd poker_hand_detection

Move the best.pt model to the /poker_hand_detection/model/ directory and rename it to playingCards.pt.

Install the required Python packages using pip:

pip install -r requirements.txt

If you have an NVIDIA GPU and want to utilize CUDA for acceleration, download and install the CUDA toolkit from the NVIDIA CUDA Downloads page.

https://developer.nvidia.com/cuda-downloads

To execute the poker_hand_detection script, run:

poker_hand_detector.py

• Ultralytics YOLOv8: YOLOv8 is a real-time object detection model developed by Ultralytics.
• Alex Bewley: For providing the SORT (Simple Online and Realtime Tracking) algorithm used for object tracking.
• Murtaza Hassan: For his comprehensive Object Detection 101 course