Skip to content

Computer vision algorithm for the recognition of the matching figure between two cards in the Dobble game.

License

Notifications You must be signed in to change notification settings

g-riverac/Computer-vision-dobble-finder

Repository files navigation

Computer Vision Project: Dobble finder

Authors:

  • Conesa Fernández, Daniel Fco
  • De Luca, Francesco
  • Díaz Pérez, Rodrigo
  • Moreno Escudero, Germán
  • Rivera Cardenas, Gabriel

Description

Computer vision algorithm for the recognition of the matching figure between two cards in the Dobble game.

The algorithm mainly uses the SIFT algorithm developed by David Lowe and the RANSAC method as a filter. The functions developed by Peter Kovesi are used to filter outliners by RANSAC.

User interface

A user interface (Dobble_compare_only.mlapp) was developed in which the comparison image can be obtained in two ways:

  • Connecting remotely via WIFI to a mobile camera with the IP_Webcam application.
  • Loads an image from the database generated during the algorithm development and performs the comparison between cards.

Once the image is loaded you should use the save button if an image has been captured by the camera or use the load image button if an image has been selected from the database. Finally, with the image loaded, you can proceed to compare the two cards.

Requeriments

  • Statistics and Machine Learning Toolbox
  • Phased Array System Toolbox

Universidad Politécnica de Madrid

MUAR, January 2021

About

Computer vision algorithm for the recognition of the matching figure between two cards in the Dobble game.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages