ChessPy: ferramenta para detecção inteligente de peças de xadrez

Abstract

Chess is one of the most researched domains in artificial intelligence history. The most powerful programs are based on sophisticated research techniques, domain-specific adaptations, and handcrafted assessment functions refined over decades by human experts. The main goal of this work is to create a robust platform for position detection during chess games by combining traditional digital image processing techniques with cutting-edge object detection algorithms. The images captured during a chess game are analyzed to determine the location of each square on the board as well as the location of each piece in play; this is then repeated for each turn so that the system can follow the game in its entirety.