Design of Automatic Game System Based on Artificial Intelligence for Three Pieces of Chess

In this paper, we design an automatic three pieces game device based on STM32F407ZGT6, aiming at realizing the accurate human-computer interaction game function. The device utilizes an air pump and X, Y gantry mechanical structure and integrates YOLOv5 deep learning model. Equipped with Emm42_V5.0 closed-loop stepper motors and MG996R servo motors, the device can precisely control the placement of the pieces in 3D space. By connecting to MaixCAM to realize piece position detection and confirmation, ensuring accurate execution of the game. the introduction of the YOLOv5 model significantly improves the accuracy and speed of piece recognition, enabling the system to respond to user actions and game changes in a timely manner. Provides a friendly user interface with physical key input. Users can notify the system to start the timer by pressing the designated button after completing a round of piece placement. Meanwhile, feedback to the user after each round of the game is provided through an indicator light, which facilitates real-time monitoring of the progress of the game.