PokerHandsRecognition

PokerHandsRecognition is a Python program that uses a camera to recognize cards from a Texas Hold'em in-real-life (IRL) game and calculates the winning or draw statistics of each hand for two players.

Index

• Features
• Files
• Usage
• Authors

Features

• Card Recognition: The program uses computer vision techniques to identify and recognize playing cards from images captured by a camera.
• Hand Analysis: It analyzes the recognized cards to determine the best possible hand combinations, such as pairs, flushes, straights, etc.
• Montecarlo Simulation: Based on the identified hands, the program calculates the winning probabilities and statistics for each player or hand combination.
• Real-time Analysis: With the camera mode, you can continuously capture and analyze card images during a Texas Hold'em game, providing instant feedback on hand strengths.

Files

• camera.py: The main Python script that runs the PokerHandsRecognition program. It captures images from the camera and performs card recognition and hand analysis.
• card.py: This script is responsible for recognizing all the cards and querying the trained model for predictions.
• keras.ipynb: A Jupyter Notebook that generates the VGG16 model. It uses images from the images/training/labeled directory for training.
• montecarlo.py: This script runs Monte Carlo simulations to determine the winning statistics.
• images/: A directory to store the captured images and the analyzed results. The original images and the analyzed images will be saved in this directory.
• images/test/: A directory containing test images that can be used to evaluate the program's performance or as examples for single photo analysis.
• images/training/labeled/: A directory containing labeled images used for training the card recognition model.
• README.md: This file that provides an overview of the repository and instructions for using the PokerHandsRecognition program.

Usage

There are two usage options for running the program:

1. Single Photo Analysis: Run the program for a single photo by providing the path to the photo as an argument. You can use the test images available in the images/test directory as examples.

   python camera.py --arg {path_to_photo}

2. Real-time Camera Analysis: Connect a camera to your computer and run the program. It will display a frame, and when you press the 'p' key, it will take a screenshot and analyze the captured image. Once the analysis is completed, it will show the final analysis and save both the original and analyzed photos in the images directory. The program will then continue waiting for another 'p' key press. To exit the program, press the 'q' key.

   python camera.py

Authors

• Álvaro Alonso Miguel

For any questions, support, or feedback, please feel free to contact me at alonso.miguel.alvaro1@gmail.com.