Skip to content

Computer vision tool to analyze Mathable game images, detect new pieces, recognize numbers, and calculate player scores automatically.

Notifications You must be signed in to change notification settings

ana-rosu/MathableScoreCalculator-ComputerVision

Repository files navigation

Mathable Score Calculator

The solution addresses the task of analyzing 200 images from 4 Mathable games, each consisting of 50 moves. During a round, players can make up to 7 moves, forming equations on the board to earn points. A round ends if no more moves are possible.

The objective is to:

  • Task 1: Identify the position of the newly added piece on the board.
  • Task 2: Recognize the number on the newly added piece.
  • Task 3: Compute the score for each round corresponding to a player by summing earned points based on valid equations formed by the new piece placement in each move of the round.

Required Libraries

To run the project, ensure the following libraries are installed:

opencv-python==4.10.0
matplotlib==3.9.2
numpy==2.0.1

You can install them using:

pip install -r requirements.txt

Usage Instructions

  1. Open the mathable.ipynb notebook.
  2. Execute all cells sequentially.

Running all tasks

  • The final cell of the notebook calls the run function with two arguments:
    1. input_folder_name: Specifies the folder containing input files.
    2. number_of_games: Specifies how many games are present in the folder.

Output

  • The results are generated in the following directory:
  evaluare/fisiere_solutie/331_Rosu_Ana_Maria

About

Computer vision tool to analyze Mathable game images, detect new pieces, recognize numbers, and calculate player scores automatically.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published